Abstract

Gaussian process models provide a flexible, non-parametric approach to modelling that sustains uncertainty about the function. However, computational demands and the joint Gaussian assumption make them inappropriate for some applications. In this talk we review low rank approximations for Gaussian processes and use stochastic process composition to create non-Gaussian processes. We illustrate the models on simple regression tasks to give a sense of how uncertainty propagates through the model. We end will demonstrations on unsupervised learning of digits and motion capture data.

.

.

Introduction [edit]

In the previous session on Gaussian processes, we introduced the Gaussian process model and the covariance function. In this session we are going to address two challenges of the Gaussian process. Firstly, we look at the computational tractability and secondly we look at extending the nature of the process beyond Gaussian.

Universe isn't as Gaussian as it Was [edit]

Figure: The cosmic microwave background is, to a very high degree of precision, a Gaussian process. The parameters of its covariance function are given by fundamental parameters of the universe, such as the amount of dark matter and mass.

=f()

Figure: What we observe today is some non-linear function of the cosmic microwave background.

Computational Complexity

Low Rank Gaussian Processes [edit]

Figure: In recent years, approximations for Gaussian process models haven't been the most fashionable approach to machine learning. Image credit: Kai Arulkumaran

Inference in a Gaussian process has computational complexity of $\bigO(\numData^3)$ and storage demands of $\bigO(\numData^2)$. This is too large for many modern data sets.

Low rank approximations allow us to work with Gaussian processes with computational complexity of $\bigO(\numData\numInducing^2)$ and storage demands of $\bigO(\numData\numInducing)$, where $\numInducing$ is a user chosen parameter.

In machine learning, low rank approximations date back to Smola and Bartlett (n.d.), Williams and Seeger (n.d.), who considered the Nyström approximation and Csató and Opper (2002);Csató (2002) who considered low rank approximations in the context of on-line learning. Selection of active points for the approximation was considered by Seeger, Williams, and Lawrence (n.d.) and Snelson and Ghahramani (n.d.) first proposed that the active set could be optimized directly. Those approaches were reviewed by Quiñonero Candela and Rasmussen (2005) under a unifying likelihood approximation perspective. General rules for deriving the maximum likelihood for these sparse approximations were given in Lawrence (n.d.).

Modern variational interpretations of these low rank approaches were first explored in Titsias (n.d.). A more modern summary which considers each of these approximations as an α-divergence is given by Bui, Yan, and Turner (2017).

Variational Compression [edit]

Inducing variables are a compression of the real observations. The basic idea is can I create a new data set that summarizes all the information in the original data set. If this data set is smaller, I've compressed the information in the original data set.

Inducing variables can be thought of as pseudo-data, indeed in Snelson and Ghahramani (n.d.) they were referred to as pseudo-points.

The only requirement for inducing variables is that they are jointly distributed as a Gaussian process with the original data. This means that they can be from the space $\mappingFunctionVector$ or a space that is related through a linear operator (see e.g. Álvarez et al. (2010)). For example we could choose to store the gradient of the function at particular points or a value from the frequency spectrum of the function (Lázaro-Gredilla, Quiñonero-Candela, and Rasmussen 2010).

Variational Compression II

Inducing variables don't only allow for the compression of the non-parameteric information into a reduced data aset but they also allow for computational scaling of the algorithms through, for example stochastic variational approaches Hensman, Fusi, and Lawrence (n.d.) or parallelization Gal, Wilk, and Rasmussen (n.d.),Dai et al. (2014), M. W. Seeger et al. (2017).

We’ve seen how we go from parametric to non-parametric. The limit implies infinite dimensional $\mappingVector$. Gaussian processes are generally non-parametric: combine data with covariance function to get model. This representation cannot be summarized by a parameter vector of a fixed size.

Parametric models have a representation that does not respond to increasing training set size. Bayesian posterior distributions over parameters contain the information about the training data, for example if we use use Bayes’ rule from training data,
$$ p\left(\mappingVector|\dataVector, \inputMatrix\right), $$
to make predictions on test data
$$ p\left(\dataScalar_*|\inputMatrix_*, \dataVector, \inputMatrix\right) = \int p\left(\dataScalar_*|\mappingVector,\inputMatrix_*\right)p\left(\mappingVector|\dataVector, \inputMatrix)\text{d}\mappingVector\right) $$
then $\mappingVector$ becomes a bottleneck for information about the training set to pass to the test set. The solution is to increase $\numBasisFunc$ so that the bottleneck is so large that it no longer presents a problem. How big is big enough for $\numBasisFunc$? Non-parametrics says $\numBasisFunc \rightarrow \infty$.

Now no longer possible to manipulate the model through the standard parametric form. However, it is possible to express parametric as GPs:
$$ \kernelScalar\left(\inputVector_i,\inputVector_j\right)=\basisFunction_:\left(\inputVector_i\right)^\top\basisFunction_:\left(\inputVector_j\right). $$
These are known as degenerate covariance matrices. Their rank is at most $\numBasisFunc$, non-parametric models have full rank covariance matrices. Most well known is the “linear kernel”,
$$ \kernelScalar(\inputVector_i, \inputVector_j) = \inputVector_i^\top\inputVector_j. $$
For non-parametrics prediction at a new point, $\mappingFunctionVector_*$, is made by conditioning on $\mappingFunctionVector$ in the joint distribution. In GPs this involves combining the training data with the covariance function and the mean function. Parametric is a special case when conditional prediction can be summarized in a fixed number of parameters. Complexity of parametric model remains fixed regardless of the size of our training data set. For a non-parametric model the required number of parameters grows with the size of the training data.

Augment Variable Space [edit]

In inducing variable approximations, we augment the variable space with a set of inducing points, $\inducingVector$. These inducing points are jointly Gaussian distributed with the points from our function, $\mappingFunctionVector$. So we have a joint Gaussian process with covariance,
$$ \begin{bmatrix} \mappingFunctionVector\\ \inducingVector \end{bmatrix} \sim \gaussianSamp{\zerosVector}{\kernelMatrix} $$
where the kernel matrix itself can be decomposed into
$$ \kernelMatrix = \begin{bmatrix} \Kff & \Kfu \\ \Kuf & \Kuu \end{bmatrix} $$

This defines a joint density between the original function points, $\mappingFunctionVector$ and our inducing points, $\inducingVector$. This can be decomposed through the product rule to give.
$$ p(\mappingFunctionVector, \inducingVector) = p(\mappingFunctionVector| \inducingVector) p(\inducingVector) $$
The Gaussian process is (typically) given by a noise corrupted form of $\mappingFunctionVector$, i.e.,
$$ \dataScalar(\inputVector) = \mappingFunction(\inputVector) + \noiseScalar, $$
which can be written probabilisticlly as,
$$ p(\dataVector) = \int p(\dataVector|\mappingFunctionVector) p(\mappingFunctionVector) \text{d}\mappingFunctionVector, $$
where for the independent case we have $p(\dataVector | \mappingFunctionVector) = \prod_{i=1}^\numData p(\dataScalar_i|\mappingFunction_i)$.

Inducing variables are like auxilliary variables in Monte Carlo algorithms. We introduce the inducing variables by augmenting this integral with an additional integral over $\inducingVector$,
$$ p(\dataVector) = \int p(\dataVector|\mappingFunctionVector) p(\mappingFunctionVector|\inducingVector) p(\inducingVector) \text{d}\inducingVector \text{d}\mappingFunctionVector. $$
Now, conceptually speaking we are going to integrate out $\mappingFunctionVector$, initially leaving $\inducingVector$ in place. This gives,
$$ p(\dataVector) = \int p(\dataVector|\inducingVector) p(\inducingVector) \text{d}\inducingVector. $$

Note the similarity between this form and our standard parametric form. If we had defined our model through standard basis functions we would have,
$$ \dataScalar(\inputVector) = \weightVector^\top\basisVector(\inputVector) + \noiseScalar $$
and the resulting probabilistic representation would be
$$ p(\dataVector) = \int p(\dataVector|\weightVector) p(\weightVector) \text{d} \weightVector $$
allowing us to predict
$$ p(\dataVector^*|\dataVector) = \int p(\dataVector^*|\weightVector) p(\weightVector|\dataVector) \text{d} \weightVector $$

The new prediction algorithm involves
$$ p(\dataVector^*|\dataVector) = \int p(\dataVector^*|\inducingVector) p(\inducingVector|\dataVector) \text{d} \inducingVector $$
but importantly the length of $\inducingVector$ is not fixed at design time like the number of parameters were. We can vary the number of inducing variables we use to give us the model capacity we require.

Unfortunately, computation of $p(\dataVector|\inducingVector)$ turns out to be intractable. As a result, we need to turn to approximations to make progress.

Variational Bound on $p(\dataVector |\inducingVector)$ [edit]

The conditional density of the data given the inducing points can be lower bounded variationally
$$ \begin{aligned} \log p(\dataVector|\inducingVector) & = \log \int p(\dataVector|\mappingFunctionVector) p(\mappingFunctionVector|\inducingVector) \text{d}\mappingFunctionVector\\ & = \int q(\mappingFunctionVector) \log \frac{p(\dataVector|\mappingFunctionVector) p(\mappingFunctionVector|\inducingVector)}{q(\mappingFunctionVector)}\text{d}\mappingFunctionVector + \KL{q(\mappingFunctionVector)}{p(\mappingFunctionVector|\dataVector, \inducingVector)}. \end{aligned} $$

The key innovation from Titsias (n.d.) was to then make a particular choice for $q(\mappingFunctionVector)$. If we set $q(\mappingFunctionVector)=p(\mappingFunctionVector|\inducingVector)$,
$$ \log p(\dataVector|\inducingVector) \geq \int p(\mappingFunctionVector|\inducingVector) \log p(\dataVector|\mappingFunctionVector)\text{d}\mappingFunctionVector. $$

$$ p(\dataVector|\inducingVector) \geq \exp \int p(\mappingFunctionVector|\inducingVector) \log p(\dataVector|\mappingFunctionVector)\text{d}\mappingFunctionVector. $$

Optimal Compression in Inducing Variables

Maximizing the lower bound minimizes the Kullback-Leibler divergence (or information gain) between our approximating density, $p(\mappingFunctionVector|\inducingVector)$ and the true posterior density, $p(\mappingFunctionVector|\dataVector, \inducingVector)$.

$$ \KL{p(\mappingFunctionVector|\inducingVector)}{p(\mappingFunctionVector|\dataVector, \inducingVector)} = \int p(\mappingFunctionVector|\inducingVector) \log \frac{p(\mappingFunctionVector|\inducingVector)}{p(\mappingFunctionVector|\dataVector, \inducingVector)}\text{d}\inducingVector $$

This bound is minimized when the information stored about $\dataVector$ is already stored in $\inducingVector$. In other words, maximizing the bound seeks an optimal compression from the information gain perspective.

For the case where $\inducingVector = \mappingFunctionVector$ the bound is exact ($\mappingFunctionVector$ d-separates $\dataVector$ from $\inducingVector$).

Choice of Inducing Variables

The quality of the resulting bound is determined by the choice of the inducing variables. You are free to choose whichever heuristics you like for the inducing variables, as long as they are drawn jointly from a valid Gaussian process, i.e. such that
$$ \begin{bmatrix} \mappingFunctionVector\\ \inducingVector \end{bmatrix} \sim \gaussianSamp{\zerosVector}{\kernelMatrix} $$
where the kernel matrix itself can be decomposed into
$$ \kernelMatrix = \begin{bmatrix} \Kff & \Kfu \\ \Kuf & \Kuu \end{bmatrix} $$
Choosing the inducing variables amounts to specifying $\Kfu$ and $\Kuu$ such that $\kernelMatrix$ remains positive definite. The typical choice is to choose $\inducingVector$ in the same domain as $\mappingFunctionVector$, associating each inducing output, $\inducingScalar_i$ with a corresponding input location $\inducingInputVector$. However, more imaginative choices are absolutely possible. In particular, if $\inducingVector$ is related to $\mappingFunctionVector$ through a linear operator (see e.g. Álvarez et al. (2010)), then valid $\Kuu$ and $\Kuf$ can be constructed. For example we could choose to store the gradient of the function at particular points or a value from the frequency spectrum of the function (Lázaro-Gredilla, Quiñonero-Candela, and Rasmussen 2010).

Variational Compression II

Inducing variables don't only allow for the compression of the non-parameteric information into a reduced data set but they also allow for computational scaling of the algorithms through, for example stochastic variational approaches(Hoffman et al. 2012; Hensman, Fusi, and Lawrence, n.d.) or parallelization (Gal, Wilk, and Rasmussen, n.d.; Dai et al. 2014; M. W. Seeger et al. 2017).

A Simple Regression Problem [edit]

Here we set up a simple one dimensional regression problem. The input locations, $\inputMatrix$, are in two separate clusters. The response variable, $\dataVector$, is sampled from a Gaussian process with an exponentiated quadratic covariance.

import numpy as np
import GPy

np.random.seed(101)

N = 50
noise_var = 0.01
X = np.zeros((50, 1))
X[:25, :] = np.linspace(0,3,25)[:,None] # First cluster of inputs/covariates
X[25:, :] = np.linspace(7,10,25)[:,None] # Second cluster of inputs/covariates

# Sample response variables from a Gaussian process with exponentiated quadratic covariance.
k = GPy.kern.RBF(1)
y = np.random.multivariate_normal(np.zeros(N),k.K(X)+np.eye(N)*np.sqrt(noise_var)).reshape(-1,1)

First we perform a full Gaussian process regression on the data. We create a GP model, m_full, and fit it to the data, plotting the resulting fit.

m_full = GPy.models.GPRegression(X,y)
_ = m_full.optimize(messages=True) # Optimize parameters of covariance function

import matplotlib.pyplot as plt
import mlai
import teaching_plots as plot 
from gp_tutorial import gpplot

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full, ax=ax, xlabel='$x$', ylabel='$y$', fontsize=20, portion=0.2)
xlim = ax.get_xlim()
ylim = ax.get_ylim()
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/sparse-demo-full-gp.svg', 
                  transparent=True, frameon=True)

Now we set up the inducing variables, u. Each inducing variable has its own associated input index, Z, which lives in the same space as $\inputMatrix$. Here we are using the true covariance function parameters to generate the fit.

kern = GPy.kern.RBF(1)
Z = np.hstack(
        (np.linspace(2.5,4.,3),
        np.linspace(7,8.5,3)))[:,None]
m = GPy.models.SparseGPRegression(X,y,kernel=kern,Z=Z)
m.noise_var = noise_var
m.inducing_inputs.constrain_fixed()
display(m)

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, ax=ax, xlabel='$x$', ylabel='$y$', fontsize=20, portion=0.2, xlim=xlim, ylim=ylim)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/sparse-demo-constrained-inducing-6-unlearned-gp.svg', 
                  transparent=True, frameon=True)

_ = m.optimize(messages=True)
display(m)

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, ax=ax, xlabel='$x$', ylabel='$y$', fontsize=20, portion=0.2, xlim=xlim, ylim=ylim)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/sparse-demo-constrained-inducing-6-learned-gp.svg', 
                  transparent=True, frameon=True)

m.randomize()
m.inducing_inputs.unconstrain()
_ = m.optimize(messages=True)

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, ax=ax, xlabel='$x$', ylabel='$y$', fontsize=20, portion=0.2,xlim=xlim, ylim=ylim)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/sparse-demo-unconstrained-inducing-6-gp.svg', 
                  transparent=True, frameon=True)

Now we will vary the number of inducing points used to form the approximation.

m.num_inducing=8
m.randomize()
M = 8
m.set_Z(np.random.rand(M,1)*12)

_ = m.optimize(messages=True)

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, ax=ax, xlabel='$x$', ylabel='$y$', fontsize=20, portion=0.2, xlim=xlim, ylim=ylim)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/sparse-demo-sparse-inducing-8-gp.svg', 
                  transparent=True, frameon=True)

And we can compare the probability of the result to the full model.

print(m.log_likelihood(), m_full.log_likelihood())

GPs for Big Data [edit]

• Let’s be explicity about storing approximate posterior of $\inducingVector$, $q(\inducingVector)$.

• Now we have
  $$p(\dataVector^*|\dataVector) = \int p(\dataVector^*| \inducingVector) q(\inducingVector | \dataVector) \text{d} \inducingVector$$
  

• Inducing variables look a lot like regular parameters.
• But: their dimensionality does not need to be set at design time.
• They can be modified arbitrarily at run time without effecting the model likelihood.

• They only effect the quality of compression and the lower bound.

• Exploit the resulting factorization ...
  $$p(\dataVector^*|\dataVector) = \int p(\dataVector^*| \inducingVector) q(\inducingVector | \dataVector) \text{d} \inducingVector$$
  
• The distribution now factorizes:
  $$p(\dataVector^*|\dataVector) = \int \prod_{i=1}^{\numData^*}p(\dataScalar^*_i| \inducingVector) q(\inducingVector | \dataVector) \text{d} \inducingVector$$
  

• This factorization can be exploited for stochastic variational inference (Hoffman et al. 2012).

Nonparametrics for Very Large Data Sets [edit]

Figure: House prices across the UK are now easily available. http://landregistry.data.gov.uk/

Figure: Zooming in on house prices across the UK's Peak District shows different areas of wealth. Could these house prices be used, e.g., as a proxy for the Index of Multiple Deprivation. Or are such indices merely intermediate indices for the real measures we're interested in?

(Hensman, Fusi, and Lawrence, n.d.)

Figure: We can now take advantage of modern variational scaling techniques such as stochastic variational inference to fit Gaussian processes to very large data sets. http://auai.org/uai2013/prints/papers/244.pdf

(Hensman, Fusi, and Lawrence, n.d.)

Figure: New approaches to scaling Gaussian processes bring the potential to model these indices and others directly. http://auai.org/uai2013/prints/papers/244.pdf

• A Unifying Framework for Gaussian Process Pseudo-Point Approximations using Power Expectation Propagation Bui, Yan, and Turner (2017)

• Deep Gaussian Processes and Variational Propagation of Uncertainty Damianou (2015)

Even in the early days of Gaussian processes in machine learning, it was understood that we were throwing something fundamental away. This is perhaps captured best by David MacKay in his 1997 NeurIPS tutorial on Gaussian processes, where he asked "Have we thrown out the baby with the bathwater?". The quote below is from his summarization paper.

According to the hype of 1987, neural networks were meant to be intelligent models which discovered features and patterns in data. Gaussian processes in contrast are simply smoothing devices. How can Gaussian processes possibly repalce neural networks? What is going on?

MacKay (n.d.)

import teaching_plots as plot

Mathematically, each layer of a neural network is given through computing the activation function, $\basisFunction(\cdot)$, contingent on the previous layer, or the inputs. In this way the activation functions, are composed to generate more complex interactions than would be possible with any single layer.
$$ \begin{align} \hiddenVector_{1} &= \basisFunction\left(\mappingMatrix_1 \inputVector\right)\\ \hiddenVector_{2} &= \basisFunction\left(\mappingMatrix_2\hiddenVector_{1}\right)\\ \hiddenVector_{3} &= \basisFunction\left(\mappingMatrix_3 \hiddenVector_{2}\right)\\ \dataVector &= \mappingVector_4 ^\top\hiddenVector_{3} \end{align} $$

Overfitting [edit]

One potential problem is that as the number of nodes in two adjacent layers increases, the number of parameters in the affine transformation between layers, $\mappingMatrix$, increases. If there are k_i − 1 nodes in one layer, and k_i nodes in the following, then that matrix contains k_ik_i − 1 parameters, when we have layer widths in the 1000s that leads to millions of parameters.

One proposed solution is known as dropout where only a sub-set of the neural network is trained at each iteration. An alternative solution would be to reparameterize $\mappingMatrix$ with its singular value decomposition.
$$ \mappingMatrix = \eigenvectorMatrix\eigenvalueMatrix\eigenvectwoMatrix^\top $$
or
$$ \mappingMatrix = \eigenvectorMatrix\eigenvectwoMatrix^\top $$
where if $\mappingMatrix \in \Re^{k_1\times k_2}$ then $\eigenvectorMatrix\in \Re^{k_1\times q}$ and $\eigenvectwoMatrix \in \Re^{k_2\times q}$, i.e. we have a low rank matrix factorization for the weights.

import teaching_plots as plot

import teaching_plots as plot

Including the low rank decomposition of $\mappingMatrix$ in the neural network, we obtain a new mathematical form. Effectively, we are adding additional latent layers, $\latentVector$, in between each of the existing hidden layers. In a neural network these are sometimes known as bottleneck layers. The network can now be written mathematically as
$$ \begin{align} \latentVector_{1} &= \eigenvectwoMatrix^\top_1 \inputVector\\ \hiddenVector_{1} &= \basisFunction\left(\eigenvectorMatrix_1 \latentVector_{1}\right)\\ \latentVector_{2} &= \eigenvectwoMatrix^\top_2 \hiddenVector_{1}\\ \hiddenVector_{2} &= \basisFunction\left(\eigenvectorMatrix_2 \latentVector_{2}\right)\\ \latentVector_{3} &= \eigenvectwoMatrix^\top_3 \hiddenVector_{2}\\ \hiddenVector_{3} &= \basisFunction\left(\eigenvectorMatrix_3 \latentVector_{3}\right)\\ \dataVector &= \mappingVector_4^\top\hiddenVector_{3}. \end{align} $$

$$ \begin{align} \latentVector_{1} &= \eigenvectwoMatrix^\top_1 \inputVector\\ \latentVector_{2} &= \eigenvectwoMatrix^\top_2 \basisFunction\left(\eigenvectorMatrix_1 \latentVector_{1}\right)\\ \latentVector_{3} &= \eigenvectwoMatrix^\top_3 \basisFunction\left(\eigenvectorMatrix_2 \latentVector_{2}\right)\\ \dataVector &= \mappingVector_4 ^\top \latentVector_{3} \end{align} $$

Now if we replace each of these neural networks with a Gaussian process. This is equivalent to taking the limit as the width of each layer goes to infinity, while appropriately scaling down the outputs.

$$ \begin{align} \latentVector_{1} &= \mappingFunctionVector_1\left(\inputVector\right)\\ \latentVector_{2} &= \mappingFunctionVector_2\left(\latentVector_{1}\right)\\ \latentVector_{3} &= \mappingFunctionVector_3\left(\latentVector_{2}\right)\\ \dataVector &= \mappingFunctionVector_4\left(\latentVector_{3}\right) \end{align} $$

Deep Learning [edit]

DeepFace [edit]

Figure: The DeepFace architecture (Taigman et al. 2014), visualized through colors to represent the functional mappings at each layer. There are 120 million parameters in the model.

The DeepFace architecture (Taigman et al. 2014) consists of layers that deal with translation and rotational invariances. These layers are followed by three locally-connected layers and two fully-connected layers. Color illustrates feature maps produced at each layer. The neural network includes more than 120 million parameters, where more than 95% come from the local and fully connected layers.

Deep Learning as Pinball [edit]

Figure: Deep learning models are composition of simple functions. We can think of a pinball machine as an analogy. Each layer of pins corresponds to one of the layers of functions in the model. Input data is represented by the location of the ball from left to right when it is dropped in from the top. Output class comes from the position of the ball as it leaves the pins at the bottom.

Sometimes deep learning models are described as being like the brain, or too complex to understand, but one analogy I find useful to help the gist of these models is to think of them as being similar to early pin ball machines.

In a deep neural network, we input a number (or numbers), whereas in pinball, we input a ball.

Think of the location of the ball on the left-right axis as a single number. Our simple pinball machine can only take one number at a time. As the ball falls through the machine, each layer of pins can be thought of as a different layer of 'neurons'. Each layer acts to move the ball from left to right.

In a pinball machine, when the ball gets to the bottom it might fall into a hole defining a score, in a neural network, that is equivalent to the decision: a classification of the input object.

An image has more than one number associated with it, so it is like playing pinball in a hyper-space.

import pods
from ipywidgets import IntSlider

pods.notebook.display_plots('pinball{sample:0>3}.svg', 
                            '../slides/diagrams',
                            sample=IntSlider(1, 1, 2, 1))

Learning involves moving all the pins to be in the correct position, so that the ball ends up in the right place when it's fallen through the machine. But moving all these pins in hyperspace can be difficult.

In a hyper-space you have to put a lot of data through the machine for to explore the positions of all the pins. Even when you feed many millions of data points through the machine, there are likely to be regions in the hyper-space where no ball has passed. When future test data passes through the machine in a new route unusual things can happen.

Adversarial examples exploit this high dimensional space. If you have access to the pinball machine, you can use gradient methods to find a position for the ball in the hyper space where the image looks like one thing, but will be classified as another.

Probabilistic methods explore more of the space by considering a range of possible paths for the ball through the machine. This helps to make them more data efficient and gives some robustness to adversarial examples.

Mathematically, a deep Gaussian process can be seen as a composite multivariate function,
$$ \mathbf{g}(\inputVector)=\mappingFunctionVector_5(\mappingFunctionVector_4(\mappingFunctionVector_3(\mappingFunctionVector_2(\mappingFunctionVector_1(\inputVector))))). $$
Or if we view it from the probabilistic perspective we can see that a deep Gaussian process is specifying a factorization of the joint density, the standard deep model takes the form of a Markov chain.

$$ p(\dataVector|\inputVector)= p(\dataVector|\mappingFunctionVector_5)p(\mappingFunctionVector_5|\mappingFunctionVector_4)p(\mappingFunctionVector_4|\mappingFunctionVector_3)p(\mappingFunctionVector_3|\mappingFunctionVector_2)p(\mappingFunctionVector_2|\mappingFunctionVector_1)p(\mappingFunctionVector_1|\inputVector) $$

Why Deep? [edit]

If the result of composing many functions together is simply another function, then why do we bother? The key point is that we can change the class of functions we are modeling by composing in this manner. A Gaussian process is specifying a prior over functions, and one with a number of elegant properties. For example, the derivative process (if it exists) of a Gaussian process is also Gaussian distributed. That makes it easy to assimilate, for example, derivative observations. But that also might raise some alarm bells. That implies that the marginal derivative distribution is also Gaussian distributed. If that's the case, then it means that functions which occasionally exhibit very large derivatives are hard to model with a Gaussian process. For example, a function with jumps in.

A one off discontinuity is easy to model with a Gaussian process, or even multiple discontinuities. They can be introduced in the mean function, or independence can be forced between two covariance functions that apply in different areas of the input space. But in these cases we will need to specify the number of discontinuities and where they occur. In otherwords we need to parameterise the discontinuities. If we do not know the number of discontinuities and don't wish to specify where they occur, i.e. if we want a non-parametric representation of discontinuities, then the standard Gaussian process doesn't help.

Stochastic Process Composition

The deep Gaussian process leads to non-Gaussian models, and non-Gaussian characteristics in the covariance function. In effect, what we are proposing is that we change the properties of the functions we are considering by composing stochastic processes. This is an approach to creating new stochastic processes from well known processes.

Additionally, we are not constrained to the formalism of the chain. For example, we can easily add single nodes emerging from some point in the depth of the chain. This allows us to combine the benefits of the graphical modelling formalism, but with a powerful framework for relating one set of variables to another, that of Gaussian processes

Difficulty for Probabilistic Approaches [edit]

The challenge for composition of probabilistic models is that you need to propagate a probability densities through non linear mappings. This allows you to create broader classes of probability density. Unfortunately it renders the resulting densities intractable.

The argument in the deep learning revolution is that deep architectures allow us to develop an abstraction of the feature set through model composition. Composing Gaussian processes is analytically intractable. To form deep Gaussian processes we use a variational approach to stack the models.

import pods

pods.notebook.display_plots('stack-gp-sample-Linear-{sample:0>1}.svg', 
                            directory='../../slides/diagrams/deepgp', sample=(0,4))

Stacked PCA [edit]

Stacking a series of linear functions simply leads to a new linear function. The use of multiple linear function merely changes the covariance of the resulting Gaussian. If
$$ \latentMatrix \sim \gaussianSamp{\zerosVector}{\eye} $$
and the ith hidden layer is a multivariate linear transformation defined by $\weightMatrix_i$,
$$ \dataMatrix = \latentMatrix\weightMatrix_1 \weightMatrix_2 \dots \weightMatrix_\numLayers $$
then the rules of multivariate Gaussians tell us that
$$ \dataMatrix \sim \gaussianSamp{\zerosVector}{\weightMatrix_\numLayers \dots \weightMatrix_1 \weightMatrix^\top_1 \dots \weightMatrix^\top_\numLayers}. $$
So the model can be replaced by one where we set $\vMatrix = \weightMatrix_\numLayers \dots \weightMatrix_2 \weightMatrix_1$. So is such a model trivial? The answer is that it depends. There are two cases in which such a model remaisn interesting. Firstly, if we make intermediate observations stemming from the chain. So, for example, if we decide that,
$$ \latentMatrix_i = \weightMatrix_i \latentMatrix_{i-1} $$
and set $\latentMatrix_{0} = \inputMatrix \sim \gaussianSamp{\zerosVector}{\eye}$, then the matrices $\weightMatrix$ inter-relate a series of jointly Gaussian observations in an interesting way, stacking the full data matrix to give
$$ \latentMatrix = \begin{bmatrix} \latentMatrix_0 \\ \latentMatrix_1 \\ \vdots \\ \latentMatrix_\numLayers \end{bmatrix} $$
we can obtain
$$\latentMatrix \sim \gaussianSamp{\zerosVector}{\begin{bmatrix} \eye & \weightMatrix^\top_1 & \weightMatrix_1^\top\weightMatrix_2^\top & \dots & \vMatrix^\top \\ \weightMatrix_1 & \weightMatrix_1 \weightMatrix_1^\top & \weightMatrix_1 \weightMatrix_1^\top \weightMatrix_2^\top & \dots & \weightMatrix_1 \vMatrix^\top \\ \weightMatrix_2 \weightMatrix_1 & \weightMatrix_2 \weightMatrix_1 \weightMatrix_1^\top & \weightMatrix_2 \weightMatrix_1 \weightMatrix_1^\top \weightMatrix_2^\top & \dots & \weightMatrix_2 \weightMatrix_1 \vMatrix^\top \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ \vMatrix & \vMatrix \weightMatrix_1^\top & \vMatrix \weightMatrix_1^\top \weightMatrix_2^\top& \dots & \vMatrix\vMatrix^\top \end{bmatrix}}$$
which is a highly structured Gaussian covariance with hierarchical dependencies between the variables $\latentMatrix_i$.

Stacked GP [edit]

pods.notebook.display_plots('stack-gp-sample-RBF-{sample:0>1}.svg', 
                            directory='../../slides/diagrams/deepgp', sample=(0,4))

Note that once the box has folded over on itself, it cannot be unfolded. So a feature that is generated near the top of the model cannot be removed further down the model.

This folding over effect happens in low dimensions. In higher dimensions it is less common.

Observation of this effect at a talk in Cambridge was one of the things that caused David Duvenaud (and collaborators) to consider the behavior of deeper Gaussian process models (Duvenaud et al. 2014).

Such folding over in the latent spaces necessarily forces the density to be non-Gaussian. Indeed, since folding-over is avoided as we increase the dimensionality of the latent spaces, such processes become more Gaussian. If we take the limit of the latent space dimensionality as it tends to infinity, the entire deep Gaussian process returns to a standard Gaussian process, with a covariance function given as a deep kernel (such as those described by Cho and Saul (2009)).

Further analysis of these deep networks has been conducted by Dunlop et al. (n.d.), who use analysis of the deep network's stationary density (treating it as a Markov chain across layers), to explore the nature of the implied process prior for a deep GP.

Both of these works, however, make constraining assumptions on the form of the Gaussian process prior at each layer (e.g. same covariance at each layer). In practice, the form of this covariance can be learnt and the densities described by the deep GP are more general than those mentioned in either of these papers.

Stacked GPs (video by David Duvenaud) [edit]

David Duvenaud also created a YouTube video to help visualize what happens as you drop through the layers of a deep GP.

GPy: A Gaussian Process Framework in Python [edit]

Figure: GPy is a BSD licensed software code base for implementing Gaussian process models in Python. It is designed for teaching and modelling. We welcome contributions which can be made through the Github repository https://github.com/SheffieldML/GPy

GPy is a BSD licensed software code base for implementing Gaussian process models in python. This allows GPs to be combined with a wide variety of software libraries.

The software itself is available on GitHub and the team welcomes contributions.

The aim for GPy is to be a probabilistic-style programming language, i.e. you specify the model rather than the algorithm. As well as a large range of covariance functions the software allows for non-Gaussian likelihoods, multivariate outputs, dimensionality reduction and approximations for larger data sets.

The GPy library can be installed via pip:

pip install GPy

This notebook depends on PyDeepGP. These libraries can be installed via pip:

pip install git+https://github.com/SheffieldML/PyDeepGP.git

Olympic Marathon Data [edit]

Gold medal times for Olympic Marathon since 1896. Marathons before 1924 didn’t have a standardised distance. Present results using pace per km. In 1904 Marathon was badly organised leading to very slow times.

Image from Wikimedia Commons http://bit.ly/16kMKHQ

The first thing we will do is load a standard data set for regression modelling. The data consists of the pace of Olympic Gold Medal Marathon winners for the Olympics from 1896 to present. First we load in the data and plot.

import numpy as np
import pods

data = pods.datasets.olympic_marathon_men()
x = data['X']
y = data['Y']

offset = y.mean()
scale = np.sqrt(y.var())

import matplotlib.pyplot as plt
import teaching_plots as plot
import mlai

xlim = (1875,2030)
ylim = (2.5, 6.5)
yhat = (y-offset)/scale

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
_ = ax.plot(x, y, 'r.',markersize=10)
ax.set_xlabel('year', fontsize=20)
ax.set_ylabel('pace min/km', fontsize=20)
ax.set_xlim(xlim)
ax.set_ylim(ylim)

mlai.write_figure(figure=fig, 
                  filename='../slides/diagrams/datasets/olympic-marathon.svg', 
                  transparent=True, 
                  frameon=True)

Things to notice about the data include the outlier in 1904, in this year, the olympics was in St Louis, USA. Organizational problems and challenges with dust kicked up by the cars following the race meant that participants got lost, and only very few participants completed.

More recent years see more consistently quick marathons.

Alan Turing [edit]

Figure: Alan Turing, in 1946 he was only 11 minutes slower than the winner of the 1948 games. Would he have won a hypothetical games held in 1946? Source: Alan Turing Internet Scrapbook.

If we had to summarise the objectives of machine learning in one word, a very good candidate for that word would be generalization. What is generalization? From a human perspective it might be summarised as the ability to take lessons learned in one domain and apply them to another domain. If we accept the definition given in the first session for machine learning,
$$ \text{data} + \text{model} \xrightarrow{\text{compute}} \text{prediction} $$
then we see that without a model we can't generalise: we only have data. Data is fine for answering very specific questions, like "Who won the Olympic Marathon in 2012?", because we have that answer stored, however, we are not given the answer to many other questions. For example, Alan Turing was a formidable marathon runner, in 1946 he ran a time 2 hours 46 minutes (just under four minutes per kilometer, faster than I and most of the other Endcliffe Park Run runners can do 5 km). What is the probability he would have won an Olympics if one had been held in 1946?

To answer this question we need to generalize, but before we formalize the concept of generalization let's introduce some formal representation of what it means to generalize in machine learning.

Our first objective will be to perform a Gaussian process fit to the data, we'll do this using the GPy software.

import GPy

m_full = GPy.models.GPRegression(x,yhat)
_ = m_full.optimize() # Optimize parameters of covariance function

The first command sets up the model, then m_full.optimize() optimizes the parameters of the covariance function and the noise level of the model. Once the fit is complete, we'll try creating some test points, and computing the output of the GP model in terms of the mean and standard deviation of the posterior functions between 1870 and 2030. We plot the mean function and the standard deviation at 200 locations. We can obtain the predictions using y_mean, y_var = m_full.predict(xt)

xt = np.linspace(1870,2030,200)[:,np.newaxis]
yt_mean, yt_var = m_full.predict(xt)
yt_sd=np.sqrt(yt_var)

Now we plot the results using the helper function in teaching_plots.

import teaching_plots as plot

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full, scale=scale, offset=offset, ax=ax, xlabel='year', ylabel='pace min/km', fontsize=20, portion=0.2)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/olympic-marathon-gp.svg', 
                  transparent=True, frameon=True)

Fit Quality

In the fit we see that the error bars (coming mainly from the noise variance) are quite large. This is likely due to the outlier point in 1904, ignoring that point we can see that a tighter fit is obtained. To see this making a version of the model, m_clean, where that point is removed.

x_clean=np.vstack((x[0:2, :], x[3:, :]))
y_clean=np.vstack((y[0:2, :], y[3:, :]))

m_clean = GPy.models.GPRegression(x_clean,y_clean)
_ = m_clean.optimize()

Deep GP Fit [edit]

Let's see if a deep Gaussian process can help here. We will construct a deep Gaussian process with one hidden layer (i.e. one Gaussian process feeding into another).

Build a Deep GP with an additional hidden layer (one dimensional) to fit the model.

import GPy
import deepgp

hidden = 1
m = deepgp.DeepGP([y.shape[1],hidden,x.shape[1]],Y=yhat, X=x, inits=['PCA','PCA'], 
                  kernels=[GPy.kern.RBF(hidden,ARD=True),
                           GPy.kern.RBF(x.shape[1],ARD=True)], # the kernels for each layer
                  num_inducing=50, back_constraint=False)

import deepgp

# Call the initalization
m.initialize()

Now optimize the model.

for layer in m.layers:
    layer.likelihood.variance.constrain_positive(warning=False)
m.optimize(messages=True,max_iters=10000)

m.staged_optimize(messages=(True,True,True))

Olympic Marathon Data Deep GP

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_sample(m, scale=scale, offset=offset, samps=10, ax=ax, 
                  xlabel='year', ylabel='pace min/km', portion = 0.225)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/olympic-marathon-deep-gp-samples.svg', 
                  transparent=True, frameon=True)

Olympic Marathon Data Deep GP

Fitted GP for each layer

Now we explore the GPs the model has used to fit each layer. First of all, we look at the hidden layer.

m.visualize(scale=scale, offset=offset, xlabel='year',
            ylabel='pace min/km',xlim=xlim, ylim=ylim,
            dataset='olympic-marathon',
            diagrams='../slides/diagrams/deepgp')

import pods

pods.notebook.display_plots('olympic-marathon-deep-gp-layer-{sample:0>1}.svg', 
                            '../slides/diagrams/deepgp', sample=(0,1))

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
m.visualize_pinball(ax=ax, scale=scale, offset=offset, points=30, portion=0.1,
                    xlabel='year', ylabel='pace km/min', vertical=True)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/olympic-marathon-deep-gp-pinball.svg', 
                  transparent=True, frameon=True)

Olympic Marathon Pinball Plot

The pinball plot shows the flow of any input ball through the deep Gaussian process. In a pinball plot a series of vertical parallel lines would indicate a purely linear function. For the olypmic marathon data we can see the first layer begins to shift from input towards the right. Note it also does so with some uncertainty (indicated by the shaded backgrounds). The second layer has less uncertainty, but bunches the inputs more strongly to the right. This input layer of uncertainty, followed by a layer that pushes inputs to the right is what gives the heteroschedastic noise.

Gene Expression Example [edit]

We now consider an example in gene expression. Gene expression is the measurement of mRNA levels expressed in cells. These mRNA levels show which genes are 'switched on' and producing data. In the example we will use a Gaussian process to determine whether a given gene is active, or we are merely observing a noise response.

Della Gatta Gene Data [edit]

• Given given expression levels in the form of a time series from Della Gatta et al. (2008).

import numpy as np
import pods

data = pods.datasets.della_gatta_TRP63_gene_expression(data_set='della_gatta',gene_number=937)

x = data['X']
y = data['Y']

offset = y.mean()
scale = np.sqrt(y.var())

import matplotlib.pyplot as plt
import teaching_plots as plot
import mlai

xlim = (-20,260)
ylim = (5, 7.5)
yhat = (y-offset)/scale

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
_ = ax.plot(x, y, 'r.',markersize=10)
ax.set_xlabel('time/min', fontsize=20)
ax.set_ylabel('expression', fontsize=20)
ax.set_xlim(xlim)
ax.set_ylim(ylim)

mlai.write_figure(figure=fig, 
                  filename='../slides/diagrams/datasets/della-gatta-gene.svg', 
                  transparent=True, 
                  frameon=True)

• Want to detect if a gene is expressed or not, fit a GP to each gene Kalaitzis and Lawrence (2011).

Figure: The example is taken from the paper "A Simple Approach to Ranking Differentially Expressed Gene Expression Time Courses through Gaussian Process Regression." Kalaitzis and Lawrence (2011).

Our first objective will be to perform a Gaussian process fit to the data, we'll do this using the GPy software.

import GPy

m_full = GPy.models.GPRegression(x,yhat)
m_full.kern.lengthscale=50
_ = m_full.optimize() # Optimize parameters of covariance function

Initialize the length scale parameter (which here actually represents a time scale of the covariance function) to a reasonable value. Default would be 1, but here we set it to 50 minutes, given points are arriving across zero to 250 minutes.

xt = np.linspace(-20,260,200)[:,np.newaxis]
yt_mean, yt_var = m_full.predict(xt)
yt_sd=np.sqrt(yt_var)

Now we plot the results using the helper function in teaching_plots.

import teaching_plots as plot

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full, scale=scale, offset=offset, ax=ax, xlabel='time/min', ylabel='expression', fontsize=20, portion=0.2)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
ax.set_title('log likelihood: {ll:.3}'.format(ll=m_full.log_likelihood()), fontsize=20)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/della-gatta-gene-gp.svg', 
                  transparent=True, frameon=True)

Now we try a model initialized with a longer length scale.

m_full2 = GPy.models.GPRegression(x,yhat)
m_full2.kern.lengthscale=2000
_ = m_full2.optimize() # Optimize parameters of covariance function

import teaching_plots as plot

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full2, scale=scale, offset=offset, ax=ax, xlabel='time/min', ylabel='expression', fontsize=20, portion=0.2)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
ax.set_title('log likelihood: {ll:.3}'.format(ll=m_full2.log_likelihood()), fontsize=20)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/della-gatta-gene-gp2.svg', 
                  transparent=True, frameon=True)

Now we try a model initialized with a lower noise.

m_full3 = GPy.models.GPRegression(x,yhat)
m_full3.kern.lengthscale=20
m_full3.likelihood.variance=0.001
_ = m_full3.optimize() # Optimize parameters of covariance function

import teaching_plots as plot

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full3, scale=scale, offset=offset, ax=ax, xlabel='time/min', ylabel='expression', fontsize=20, portion=0.2)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
ax.set_title('log likelihood: {ll:.3}'.format(ll=m_full3.log_likelihood()), fontsize=20)
mlai.write_figure(figure=fig,
                  filename='../slides/diagrams/gp/della-gatta-gene-gp3.svg', 
                  transparent=True, frameon=True)

layers = [y.shape[1], 1,x.shape[1]]
inits = ['PCA']*(len(layers)-1)
kernels = []
for i in layers[1:]:
    kernels += [GPy.kern.RBF(i)]
m = deepgp.DeepGP(layers,Y=yhat, X=x, 
                  inits=inits, 
                  kernels=kernels, # the kernels for each layer
                  num_inducing=20, back_constraint=False)

m.initialize()
m.staged_optimize()

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, scale=scale, offset=offset, ax=ax, fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(filename='../slides/diagrams/deepgp/della-gatta-gene-deep-gp.svg', 
            transparent=True, frameon=True)

Della Gatta Gene Data Deep GP

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_sample(m, scale=scale, offset=offset, samps=10, ax=ax, portion = 0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/della-gatta-gene-deep-gp-samples.svg', 
                  transparent=True, frameon=True)

Della Gatta Gene Data Deep GP

m.visualize(offset=offset, scale=scale, xlim=xlim, ylim=ylim,
            dataset='della-gatta-gene',
            diagrams='../slides/diagrams/deepgp')

Della Gatta Gene Data Latent 1

Della Gatta Gene Data Latent 2

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
m.visualize_pinball(offset=offset, ax=ax, scale=scale, xlim=xlim, ylim=ylim, portion=0.1, points=50)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/della-gatta-gene-deep-gp-pinball.svg', 
                  transparent=True, frameon=True, ax=ax)

TP53 Gene Pinball Plot

Step Function [edit]

Next we consider a simple step function data set.

num_low=25
num_high=25
gap = -.1
noise=0.0001
x = np.vstack((np.linspace(-1, -gap/2.0, num_low)[:, np.newaxis],
              np.linspace(gap/2.0, 1, num_high)[:, np.newaxis]))
y = np.vstack((np.zeros((num_low, 1)), np.ones((num_high,1))))
scale = np.sqrt(y.var())
offset = y.mean()
yhat = (y-offset)/scale

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
_ = ax.plot(x, y, 'r.',markersize=10)
_ = ax.set_xlabel('$x$', fontsize=20)
_ = ax.set_ylabel('$y$', fontsize=20)
xlim = (-2, 2)
ylim = (-0.6, 1.6)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(figure=fig, filename='../../slides/diagrams/datasets/step-function.svg', 
            transparent=True, frameon=True)

Step Function Data

Step Function Data GP

We can fit a Gaussian process to the step function data using GPy as follows.

m_full = GPy.models.GPRegression(x,yhat)
_ = m_full.optimize() # Optimize parameters of covariance function

Where GPy.models.GPRegression() gives us a standard GP regression model with exponentiated quadratic covariance function.

The model is optimized using m_full.optimize() which calls an L-BGFS gradient based solver in python.

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full, scale=scale, offset=offset, ax=ax, fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)

mlai.write_figure(figure=fig,filename='../slides/diagrams/gp/step-function-gp.svg', 
            transparent=True, frameon=True)

The resulting fit to the step function data shows some challenges. In particular, the over smoothing at the discontinuity. If we know how many discontinuities there are, we can parameterize them in the step function. But by doing this, we form a semi-parametric model. The parameters indicate how many discontinuities are, and where they are. They can be optimized as part of the model fit. But if new, unforeseen, discontinuities arise when the model is being deployed in practice, these won't be accounted for in the predictions.

Step Function Data Deep GP

First we initialize a deep Gaussian process with three latent layers (four layers total). Within each layer we create a GP with an exponentiated quadratic covariance (GPy.kern.RBF).

At each layer we use 20 inducing points for the variational approximation.

layers = [y.shape[1], 1, 1, 1,x.shape[1]]
inits = ['PCA']*(len(layers)-1)
kernels = []
for i in layers[1:]:
    kernels += [GPy.kern.RBF(i)]
    
m = deepgp.DeepGP(layers,Y=yhat, X=x, 
                  inits=inits, 
                  kernels=kernels, # the kernels for each layer
                  num_inducing=20, back_constraint=False)

Once the model is constructed we initialize the parameters, and perform the staged optimization which starts by optimizing variational parameters with a low noise and proceeds to optimize the whole model.

m.initialize()
m.staged_optimize()

We plot the output of the deep Gaussian process fitted to the stpe data as follows.

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, scale=scale, offset=offset, ax=ax, fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(filename='../slides/diagrams/deepgp/step-function-deep-gp.svg', 
            transparent=True, frameon=True)

The deep Gaussian process does a much better job of fitting the data. It handles the discontinuity easily, and error bars drop to smaller values in the regions of data.

Step Function Data Deep GP

The samples of the model can be plotted with the helper function from teaching_plots.py, model_sample

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)

plot.model_sample(m, scale=scale, offset=offset, samps=10, ax=ax, portion = 0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/step-function-deep-gp-samples.svg', 
                  transparent=True, frameon=True)

The samples from the model show that the error bars, which are informative for Gaussian outputs, are less informative for this model. They make clear that the data points lie, in output mainly at 0 or 1, or occasionally in between.

The visualize code allows us to inspect the intermediate layers in the deep GP model to understand how it has reconstructed the step function.

m.visualize(offset=offset, scale=scale, xlim=xlim, ylim=ylim,
            dataset='step-function',
            diagrams='../slides/diagrams/deepgp')

A pinball plot can be created for the resulting model to understand how the input is being translated to the output across the different layers.

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
m.visualize_pinball(offset=offset, ax=ax, scale=scale, xlim=xlim, ylim=ylim, portion=0.1, points=50)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/step-function-deep-gp-pinball.svg', 
                  transparent=True, frameon=True, ax=ax)

import pods

data = pods.datasets.mcycle()
x = data['X']
y = data['Y']
scale=np.sqrt(y.var())
offset=y.mean()
yhat = (y - offset)/scale

fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
_ = ax.plot(x, y, 'r.',markersize=10)
_ = ax.set_xlabel('time', fontsize=20)
_ = ax.set_ylabel('acceleration', fontsize=20)
xlim = (-20, 80)
ylim = (-175, 125)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
mlai.write_figure(filename='../slides/diagrams/datasets/motorcycle-helmet.svg', 
            transparent=True, frameon=True)

Motorcycle Helmet Data [edit]

m_full = GPy.models.GPRegression(x,yhat)
_ = m_full.optimize() # Optimize parameters of covariance function

Motorcycle Helmet Data GP

import deepgp

layers = [y.shape[1], 1, x.shape[1]]
inits = ['PCA']*(len(layers)-1)
kernels = []
for i in layers[1:]:
    kernels += [GPy.kern.RBF(i)]
m = deepgp.DeepGP(layers,Y=yhat, X=x, 
                  inits=inits, 
                  kernels=kernels, # the kernels for each layer
                  num_inducing=20, back_constraint=False)



m.initialize()

m.staged_optimize(iters=(1000,1000,10000), messages=(True, True, True))

import teaching_plots as plot
import mlai

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, scale=scale, offset=offset, ax=ax, xlabel='time', ylabel='acceleration/$g$', fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(filename='../slides/diagrams/deepgp/motorcycle-helmet-deep-gp.svg', 
            transparent=True, frameon=True)

Motorcycle Helmet Data Deep GP

import teaching_plots as plot
import mlai

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_sample(m, scale=scale, offset=offset, samps=10, ax=ax, xlabel='time', ylabel='acceleration/$g$', portion = 0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)

mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/motorcycle-helmet-deep-gp-samples.svg', 
                  transparent=True, frameon=True)

Motorcycle Helmet Data Deep GP

m.visualize(xlim=xlim, ylim=ylim, scale=scale,offset=offset, 
            xlabel="time", ylabel="acceleration/$g$", portion=0.5,
            dataset='motorcycle-helmet',
            diagrams='../slides/diagrams/deepgp')

Motorcycle Helmet Data Latent 1

Motorcycle Helmet Data Latent 2

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
m.visualize_pinball(ax=ax, xlabel='time', ylabel='acceleration/g', 
                    points=50, scale=scale, offset=offset, portion=0.1)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/motorcycle-helmet-deep-gp-pinball.svg', 
                  transparent=True, frameon=True)

Motorcycle Helmet Pinball Plot

Robot Wireless Data [edit]

The robot wireless data is taken from an experiment run by Brian Ferris at University of Washington. It consists of the measurements of WiFi access point signal strengths as Brian walked in a loop.

data=pods.datasets.robot_wireless()

x = np.linspace(0,1,215)[:, np.newaxis]
y = data['Y']
offset = y.mean()
scale = np.sqrt(y.var())
yhat = (y-offset)/scale

The ground truth is recorded in the data, the actual loop is given in the plot below.

fig, ax = plt.subplots(figsize=plot.big_figsize)
plt.plot(data['X'][:, 1], data['X'][:, 2], 'r.', markersize=5)
ax.set_xlabel('x position', fontsize=20)
ax.set_ylabel('y position', fontsize=20)
mlai.write_figure(figure=fig, filename='../../slides/diagrams/datasets/robot-wireless-ground-truth.svg', transparent=True, frameon=True)

Robot Wireless Ground Truth

We will ignore this ground truth in making our predictions, but see if the model can recover something similar in one of the latent layers.

output_dim=1
xlim = (-0.3, 1.3)
fig, ax = plt.subplots(figsize=plot.big_wide_figsize)
_ = ax.plot(x.flatten(), y[:, output_dim], 
            'r.', markersize=5)

ax.set_xlabel('time', fontsize=20)
ax.set_ylabel('signal strength', fontsize=20)
xlim = (-0.2, 1.2)
ylim = (-0.6, 2.0)
ax.set_xlim(xlim)
ax.set_ylim(ylim)

mlai.write_figure(figure=fig, filename='../slides/diagrams/datasets/robot-wireless-dim-' + str(output_dim) + '.svg', 
            transparent=True, frameon=True)

Robot WiFi Data

Perform a Gaussian process fit on the data using GPy.

m_full = GPy.models.GPRegression(x,yhat)
_ = m_full.optimize() # Optimize parameters of covariance function

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m_full, output_dim=output_dim, scale=scale, offset=offset, ax=ax, 
                  xlabel='time', ylabel='signal strength', fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(filename='../slides/diagrams/gp/robot-wireless-gp-dim-' + str(output_dim)+ '.svg', 
            transparent=True, frameon=True)

Robot WiFi Data GP

}{Gaussian process fit to the Robot Wireless dimension 1.}{robot-wireless-gp-dim-1}

layers = [y.shape[1], 10, 5, 2, 2, x.shape[1]]
inits = ['PCA']*(len(layers)-1)
kernels = []
for i in layers[1:]:
    kernels += [GPy.kern.RBF(i, ARD=True)]

m = deepgp.DeepGP(layers,Y=y, X=x, inits=inits, 
                  kernels=kernels,
                  num_inducing=50, back_constraint=False)
m.initialize()

m.staged_optimize(messages=(True,True,True))

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_output(m, output_dim=output_dim, scale=scale, offset=offset, ax=ax, 
                  xlabel='time', ylabel='signal strength', fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/robot-wireless-deep-gp-dim-' + str(output_dim)+ '.svg', 
                  transparent=True, frameon=True)

Robot WiFi Data Deep GP

fig, ax=plt.subplots(figsize=plot.big_wide_figsize)
plot.model_sample(m, output_dim=output_dim, scale=scale, offset=offset, samps=10, ax=ax,
                  xlabel='time', ylabel='signal strength', fontsize=20, portion=0.5)
ax.set_ylim(ylim)
ax.set_xlim(xlim)
mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/robot-wireless-deep-gp-samples-dim-' + str(output_dim)+ '.svg', 
                  transparent=True, frameon=True)

Robot WiFi Data Deep GP

fig, ax = plt.subplots(figsize=plot.big_figsize)
ax.plot(m.layers[-2].latent_space.mean[:, 0], 
        m.layers[-2].latent_space.mean[:, 1], 
        'r.-', markersize=5)

ax.set_xlabel('latent dimension 1', fontsize=20)
ax.set_ylabel('latent dimension 2', fontsize=20)

mlai.write_figure(figure=fig, filename='../slides/diagrams/deepgp/robot-wireless-latent-space.svg', 
            transparent=True, frameon=True)

Motion Capture [edit]

• ‘High five’ data.
• Model learns structure between two interacting subjects.

Shared LVM

Fitting a GP to the USPS Digits Data [edit]

Thanks to: Zhenwen Dai and Neil D. Lawrence

We now look at the deep Gaussian processes' capacity to perform unsupervised learning.

We will look at a sub-sample of the MNIST digit data set.

First load in the MNIST data set from scikit learn. This can take a little while because it's large to download.

from sklearn.datasets import fetch_mldata

mnist = fetch_mldata('MNIST original')

Sub-sample the dataset to make the training faster.

import numpy as np

np.random.seed(0)
digits = [0,1,2,3,4]
N_per_digit = 100
Y = []
labels = []
for d in digits:
    imgs = mnist['data'][mnist['target']==d]
    Y.append(imgs[np.random.permutation(imgs.shape[0])][:N_per_digit])
    labels.append(np.ones(N_per_digit)*d)
Y = np.vstack(Y).astype(np.float64)
labels = np.hstack(labels)
Y /= 255.

Fit a Deep GP

We're going to fit a Deep Gaussian process model to the MNIST data with two hidden layers. Each of the two Gaussian processes (one from the first hidden layer to the second, one from the second hidden layer to the data) has an exponentiated quadratic covariance.

import deepgp
import GPy

num_latent = 2
num_hidden_2 = 5
m = deepgp.DeepGP([Y.shape[1],num_hidden_2,num_latent],
                  Y,
                  kernels=[GPy.kern.RBF(num_hidden_2,ARD=True), 
                           GPy.kern.RBF(num_latent,ARD=False)], 
                  num_inducing=50, back_constraint=False, 
                  encoder_dims=[[200],[200]])

Initialization

Just like deep neural networks, there are some tricks to intitializing these models. The tricks we use here include some early training of the model with model parameters constrained. This gives the variational inducing parameters some scope to tighten the bound for the case where the noise variance is small and the variances of the Gaussian processes are around 1.

m.obslayer.likelihood.variance[:] = Y.var()*0.01
for layer in m.layers:
    layer.kern.variance.fix(warning=False)
    layer.likelihood.variance.fix(warning=False)

We now we optimize for a hundred iterations with the constrained model.

m.optimize(messages=False,max_iters=100)

Now we remove the fixed constraint on the kernel variance parameters, but keep the noise output constrained, and run for a further 100 iterations.

for layer in m.layers:
    layer.kern.variance.constrain_positive(warning=False)
m.optimize(messages=False,max_iters=100)

Finally we unconstrain the layer likelihoods and allow the full model to be trained for 1000 iterations.

for layer in m.layers:
    layer.likelihood.variance.constrain_positive(warning=False)
m.optimize(messages=True,max_iters=10000)

Visualize the latent space of the top layer

Now the model is trained, let's plot the mean of the posterior distributions in the top latent layer of the model.

import matplotlib.pyplot as plt
from matplotlib import rc
import teaching_plots as plot
import mlai

rc("font", **{'family':'sans-serif','sans-serif':['Helvetica'],'size':20})
fig, ax = plt.subplots(figsize=plot.big_figsize)
for d in digits:
    ax.plot(m.layer_1.X.mean[labels==d,0],m.layer_1.X.mean[labels==d,1],'.',label=str(d))
_ = plt.legend()
mlai.write_figure(figure=fig, filename="../slides/diagrams/deepgp/usps-digits-latent.svg", transparent=True)

Visualize the latent space of the intermediate layer

We can also visualize dimensions of the intermediate layer. First the lengthscale of those dimensions is given by

m.obslayer.kern.lengthscale

import matplotlib.pyplot as plt
import mlai

fig, ax = plt.subplots(figsize=plot.big_figsize)
for i in range(5):
    for j in range(i):
        dims=[i, j]
        ax.cla()
        for d in digits:
            ax.plot(m.obslayer.X.mean[labels==d,dims[0]],
                 m.obslayer.X.mean[labels==d,dims[1]],
                 '.', label=str(d))
        plt.legend()
        plt.xlabel('dimension ' + str(dims[0]))
        plt.ylabel('dimension ' + str(dims[1]))
        mlai.write_figure(figure=fig, filename="../slides/diagrams/deepgp/usps-digits-hidden-" + str(dims[0]) + '-' + str(dims[1]) + '.svg', transparent=True)

Generate From Model

Now we can take a look at a sample from the model, by drawing a Gaussian random sample in the latent space and propagating it through the model.

rows = 10
cols = 20
t=np.linspace(-1, 1, rows*cols)[:, None]
kern = GPy.kern.RBF(1,lengthscale=0.05)
cov = kern.K(t, t)
x = np.random.multivariate_normal(np.zeros(rows*cols), cov, num_latent).T

import matplotlib.pyplot as plt
import mlai

yt = m.predict(x)
fig, axs = plt.subplots(rows,cols,figsize=(10,6))
for i in range(rows):
    for j in range(cols):
        #v = np.random.normal(loc=yt[0][i*cols+j, :], scale=np.sqrt(yt[1][i*cols+j, :]))
        v = yt[0][i*cols+j, :]
        axs[i,j].imshow(v.reshape(28,28), 
                        cmap='gray', interpolation='none',
                        aspect='equal')
        axs[i,j].set_axis_off()
mlai.write_figure(figure=fig, filename="../slides/diagrams/deepgp/digit-samples-deep-gp.svg", transparent=True)

Deep Health [edit]

• Gaussian process based nonlinear latent structure discovery in multivariate spike train data Wu et al. (2017)
• Doubly Stochastic Variational Inference for Deep Gaussian Processes Salimbeni and Deisenroth (2017)
• Deep Multi-task Gaussian Processes for Survival Analysis with Competing Risks Alaa and van der Schaar (2017)

• Counterfactual Gaussian Processes for Reliable Decision-making and What-if Reasoning Schulam and Saria (2017)

• Deep Survival Analysis Ranganath et al. (2016)
• Recurrent Gaussian Processes Mattos et al. (2015)

• Gaussian Process Based Approaches for Survival Analysis A. D. Saul (2016)

Emukit Playground [edit]

Emukit playground is a software toolkit for exploring the use of statistical emulation as a tool. It was built by Adam Hirst, during his software engineering internship at Amazon and supervised by Cliff McCollum.

Figure: Emukit playground is a tutorial for understanding the simulation/emulation relationship. https://amzn.github.io/emukit-playground/

Figure: Tutorial on Bayesian optimization of the number of taxis deployed from Emukit playground. https://amzn.github.io/emukit-playground/#!/learn/bayesian_optimization

You can explore Bayesian optimization of a taxi simulation.

Uncertainty Quantification [edit]

Uncertainty quantification (UQ) is the science of quantitative characterization and reduction of uncertainties in both computational and real world applications. It tries to determine how likely certain outcomes are if some aspects of the system are not exactly known.

We will to illustrate different concepts of Uncertainty Quantification (UQ) and the role that Gaussian processes play in this field. Based on a simple simulator of a car moving between a valley and a mountain, we are going to illustrate the following concepts:

• Systems emulation. Many real world decisions are based on simulations that can be computationally very demanding. We will show how simulators can be replaced by emulators: Gaussian process models fitted on a few simulations that can be used to replace the simulator. Emulators are cheap to compute, fast to run, and always provide ways to quantify the uncertainty of how precise they are compared the original simulator.

• Emulators in optimization problems. We will show how emulators can be used to optimize black-box functions that are expensive to evaluate. This field is also called Bayesian Optimization and has gained an increasing relevance in machine learning as emulators can be used to optimize computer simulations (and machine learning algorithms) quite efficiently.

• Multi-fidelity emulation methods. In many scenarios we have simulators of different quality about the same measure of interest. In these cases the goal is to merge all sources of information under the same model so the final emulator is cheaper and more accurate than an emulator fitted only using data from the most accurate and expensive simulator.

Mountain Car Simulator [edit]

To illustrate the above mentioned concepts we we use the mountain car simulator. This simulator is widely used in machine learning to test reinforcement learning algorithms. The goal is to define a control policy on a car whose objective is to climb a mountain. Graphically, the problem looks as follows:

Figure: The mountain car simulation from the Open AI gym.

The goal is to define a sequence of actions (push the car right or left with certain intensity) to make the car reach the flag after a number T of time steps.

At each time step t, the car is characterized by a vector $\inputVector_{t} = (p_t,v_t)$ of states which are respectively the the position and velocity of the car at time t. For a sequence of states (an episode), the dynamics of the car is given by

$$\inputVector_{t+1} = \mappingFunction(\inputVector_{t},\textbf{u}_{t})$$

where u_t is the value of an action force, which in this example corresponds to push car to the left (negative value) or to the right (positive value). The actions across a full episode are represented in a policy $\textbf{u}_{t} = \pi(\inputVector_{t},\theta)$ that acts according to the current state of the car and some parameters θ. In the following examples we will assume that the policy is linear which allows us to write $\pi(\inputVector_{t},\theta)$ as

$$\pi(\inputVector,\theta)= \theta_0 + \theta_p p + \theta_vv.$$

For t = 1, …, T now given some initial state $\inputVector_{0}$ and some some values of each u_t, we can simulate the full dynamics of the car for a full episode using Gym. The values of u_t are fully determined by the parameters of the linear controller.

After each episode of length T is complete, a reward function R_T(θ) is computed. In the mountain car example the reward is computed as 100 for reaching the target of the hill on the right hand side, minus the squared sum of actions (a real negative to push to the left and a real positive to push to the right) from start to goal. Note that our reward depend on θ as we make it dependent on the parameters of the linear controller.

Emulate the Mountain Car

import gym

env = gym.make('MountainCarContinuous-v0')

Our goal in this section is to find the parameters θ of the linear controller such that

θ^* = argmax_θR_T(θ).

In this section, we directly use Bayesian optimization to solve this problem. We will use GPyOpt so we first define the objective function:

import mountain_car as mc
import GPyOpt

obj_func = lambda x: mc.run_simulation(env, x)[0]
objective = GPyOpt.core.task.SingleObjective(obj_func)

For each set of parameter values of the linear controller we can run an episode of the simulator (that we fix to have a horizon of T = 500) to generate the reward. Using as input the parameters of the controller and as outputs the rewards we can build a Gaussian process emulator of the reward.

We start defining the input space, which is three-dimensional:

## --- We define the input space of the emulator

space= [{'name':'postion_parameter', 'type':'continuous', 'domain':(-1.2, +1)},
        {'name':'velocity_parameter', 'type':'continuous', 'domain':(-1/0.07, +1/0.07)},
        {'name':'constant', 'type':'continuous', 'domain':(-1, +1)}]

design_space = GPyOpt.Design_space(space=space)

Now we initizialize a Gaussian process emulator.

model = GPyOpt.models.GPModel(optimize_restarts=5, verbose=False, exact_feval=True, ARD=True)

In Bayesian optimization an acquisition function is used to balance exploration and exploitation to evaluate new locations close to the optimum of the objective. In this notebook we select the expected improvement (EI). For further details have a look to the review paper of Shahriari et al (2015).

aquisition_optimizer = GPyOpt.optimization.AcquisitionOptimizer(design_space)
acquisition = GPyOpt.acquisitions.AcquisitionEI(model, design_space, optimizer=aquisition_optimizer)
evaluator = GPyOpt.core.evaluators.Sequential(acquisition) # Collect points sequentially, no parallelization.

To initalize the model we start sampling some initial points (25) for the linear controler randomly.

from GPyOpt.experiment_design.random_design import RandomDesign

n_initial_points = 25
random_design = RandomDesign(design_space)
initial_design = random_design.get_samples(n_initial_points)

Before we start any optimization, lets have a look to the behavior of the car with the first of these initial points that we have selected randomly.

import numpy as np

random_controller = initial_design[0,:]
_, _, _, frames = mc.run_simulation(env, np.atleast_2d(random_controller), render=True)
anim=mc.animate_frames(frames, 'Random linear controller')

from IPython.core.display import HTML

HTML(anim.to_jshtml())

As we can see the random linear controller does not manage to push the car to the top of the mountain. Now, let's optimize the regret using Bayesian optimization and the emulator for the reward. We try 50 new parameters chosen by the EI.

max_iter = 50
bo = GPyOpt.methods.ModularBayesianOptimization(model, design_space, objective, acquisition, evaluator, initial_design)
bo.run_optimization(max_iter = max_iter )

Now we visualize the result for the best controller that we have found with Bayesian optimization.

_, _, _, frames = mc.run_simulation(env, np.atleast_2d(bo.x_opt), render=True)
anim=mc.animate_frames(frames, 'Best controller after 50 iterations of Bayesian optimization')

HTML(anim.to_jshtml())

he car can now make it to the top of the mountain! Emulating the reward function and using the EI helped as to find a linear controller that solves the problem.

Data Efficient Emulation [edit]

In the previous section we solved the mountain car problem by directly emulating the reward but no considerations about the dynamics $\inputVector_{t+1} = \mappingFunction(\inputVector_{t},\textbf{u}_{t})$ of the system were made. Note that we had to run 75 episodes of 500 steps each to solve the problem, which required to call the simulator 500 × 75 = 37500 times. In this section we will show how it is possible to reduce this number by building an emulator for f that can later be used to directly optimize the control.

The inputs of the model for the dynamics are the velocity, the position and the value of the control so create this space accordingly.

import gym

env = gym.make('MountainCarContinuous-v0')

import GPyOpt

space_dynamics = [{'name':'position', 'type':'continuous', 'domain':[-1.2, +0.6]},
                  {'name':'velocity', 'type':'continuous', 'domain':[-0.07, +0.07]},
                  {'name':'action', 'type':'continuous', 'domain':[-1, +1]}]
design_space_dynamics = GPyOpt.Design_space(space=space_dynamics)

The outputs are the velocity and the position. Indeed our model will capture the change in position and velocity on time. That is, we will model

Δv_t + 1 = v_t + 1 − v_t

Δx_t + 1 = p_t + 1 − p_t

with Gaussian processes with prior mean v_t and p_t respectively. As a covariance function, we use a Matern52. We need therefore two models to capture the full dynamics of the system.

position_model = GPyOpt.models.GPModel(optimize_restarts=5, verbose=False, exact_feval=True, ARD=True)
velocity_model = GPyOpt.models.GPModel(optimize_restarts=5, verbose=False, exact_feval=True, ARD=True)

Next, we sample some input parameters and use the simulator to compute the outputs. Note that in this case we are not running the full episodes, we are just using the simulator to compute $\inputVector_{t+1}$ given $\inputVector_{t}$ and u_t.

import numpy as np
from GPyOpt.experiment_design.random_design import RandomDesign
import mountain_car as mc

### --- Random locations of the inputs
n_initial_points = 500
random_design_dynamics = RandomDesign(design_space_dynamics)
initial_design_dynamics = random_design_dynamics.get_samples(n_initial_points)

### --- Simulation of the (normalized) outputs
y = np.zeros((initial_design_dynamics.shape[0], 2))
for i in range(initial_design_dynamics.shape[0]):
    y[i, :] = mc.simulation(initial_design_dynamics[i, :])

# Normalize the data from the simulation
y_normalisation = np.std(y, axis=0)
y_normalised = y/y_normalisation

In general we might use much smarter strategies to design our emulation of the simulator. For example, we could use the variance of the predictive distributions of the models to collect points using uncertainty sampling, which will give us a better coverage of the space. For simplicity, we move ahead with the 500 randomly selected points.

Now that we have a data set, we can update the emulators for the location and the velocity.

position_model.updateModel(initial_design_dynamics, y[:, [0]], None, None)
velocity_model.updateModel(initial_design_dynamics, y[:, [1]], None, None)

We can now have a look to how the emulator and the simulator match. First, we show a contour plot of the car aceleration for each pair of can position and velocity. You can use the bar bellow to play with the values of the controler to compare the emulator and the simulator.

from IPython.html.widgets import interact

We can see how the emulator is doing a fairly good job approximating the simulator. On the edges, however, it struggles to captures the dynamics of the simulator.

Given some input parameters of the linear controlling, how do the dynamics of the emulator and simulator match? In the following figure we show the position and velocity of the car for the 500 time steps of an episode in which the parameters of the linear controller have been fixed beforehand. The value of the input control is also shown.

controller_gains = np.atleast_2d([0, .6, 1])  # change the valus of the linear controller to observe the trayectories.

We now make explicit use of the emulator, using it to replace the simulator and optimize the linear controller. Note that in this optimization, we don't need to query the simulator anymore as we can reproduce the full dynamics of an episode using the emulator. For illustrative purposes, in this example we fix the initial location of the car.

We define the objective reward function in terms of the simulator.

### --- Optimize control parameters with emulator
car_initial_location = np.asarray([-0.58912799, 0]) 

### --- Reward objective function using the emulator
obj_func_emulator = lambda x: mc.run_emulation([position_model, velocity_model], x, car_initial_location)[0]
objective_emulator = GPyOpt.core.task.SingleObjective(obj_func_emulator)

And as before, we use Bayesian optimization to find the best possible linear controller.

### --- Elements of the optimization that will use the multi-fidelity emulator
model = GPyOpt.models.GPModel(optimize_restarts=5, verbose=False, exact_feval=True, ARD=True)

The design space is the three continuous variables that make up the linear controller.

space= [{'name':'linear_1', 'type':'continuous', 'domain':(-1/1.2, +1)},
        {'name':'linear_2', 'type':'continuous', 'domain':(-1/0.07, +1/0.07)},
        {'name':'constant', 'type':'continuous', 'domain':(-1, +1)}]

design_space         = GPyOpt.Design_space(space=space)
aquisition_optimizer = GPyOpt.optimization.AcquisitionOptimizer(design_space)

random_design = RandomDesign(design_space)
initial_design = random_design.get_samples(25)

We set the acquisition function to be expected improvement using GPyOpt.

acquisition          = GPyOpt.acquisitions.AcquisitionEI(model, design_space, optimizer=aquisition_optimizer)
evaluator            = GPyOpt.core.evaluators.Sequential(acquisition)

bo_emulator = GPyOpt.methods.ModularBayesianOptimization(model, design_space, objective_emulator, acquisition, evaluator, initial_design)
bo_emulator.run_optimization(max_iter=50)

_, _, _, frames = mc.run_simulation(env, np.atleast_2d(bo_emulator.x_opt), render=True)
anim=mc.animate_frames(frames, 'Best controller using the emulator of the dynamics')

from IPython.core.display import HTML

HTML(anim.to_jshtml())

And the problem is again solved, but in this case we have replaced the simulator of the car dynamics by a Gaussian process emulator that we learned by calling the simulator only 500 times. Compared to the 37500 calls that we needed when applying Bayesian optimization directly on the simulator this is a great gain.

Multi-Fidelity Emulation [edit]

In some scenarios we have simulators of the same environment that have different fidelities, that is that reflect with different level of accuracy the dynamics of the real world. Running simulations of the different fidelities also have a different cost: hight fidelity simulations are more expensive the cheaper ones. If we have access to these simulators we can combine high and low fidelity simulations under the same model.

So let's assume that we have two simulators of the mountain car dynamics, one of high fidelity (the one we have used) and another one of low fidelity. The traditional approach to this form of multi-fidelity emulation is to assume that

$$\mappingFunction_i\left(\inputVector\right) = \rho\mappingFunction_{i-1}\left(\inputVector\right) + \delta_i\left(\inputVector \right)$$

where $\mappingFunction_{i-1}\left(\inputVector\right)$ is a low fidelity simulation of the problem of interest and $\mappingFunction_i\left(\inputVector\right)$ is a higher fidelity simulation. The function $\delta_i\left(\inputVector \right)$ represents the difference between the lower and higher fidelity simulation, which is considered additive. The additive form of this covariance means that if $\mappingFunction_{0}\left(\inputVector\right)$ and $\left\{\delta_i\left(\inputVector \right)\right\}_{i=1}^m$ are all Gaussian processes, then the process over all fidelities of simuation will be a joint Gaussian process.

But with Deep Gaussian processes we can consider the form

$$\mappingFunction_i\left(\inputVector\right) = \mappingFunctionTwo_{i}\left(\mappingFunction_{i-1}\left(\inputVector\right)\right) + \delta_i\left(\inputVector \right),$$

where the low fidelity representation is non linearly transformed by $\mappingFunctionTwo(\cdot)$ before use in the process. This is the approach taken in Perdikaris et al. (2017). But once we accept that these models can be composed, a highly flexible framework can emerge. A key point is that the data enters the model at different levels, and represents different aspects. For example these correspond to the two fidelities of the mountain car simulator.

We start by sampling both of them at 250 random input locations.

import gym

env = gym.make('MountainCarContinuous-v0')

import GPyOpt

### --- Collect points from low and high fidelity simulator --- ###

space = GPyOpt.Design_space([
        {'name':'position', 'type':'continuous', 'domain':(-1.2, +1)},
        {'name':'velocity', 'type':'continuous', 'domain':(-0.07, +0.07)},
        {'name':'action', 'type':'continuous', 'domain':(-1, +1)}])

n_points = 250
random_design = GPyOpt.experiment_design.RandomDesign(space)
x_random = random_design.get_samples(n_points)

Next, we evaluate the high and low fidelity simualtors at those locations.

import numpy as np
import mountain_car as mc

d_position_hf = np.zeros((n_points, 1))
d_velocity_hf = np.zeros((n_points, 1))
d_position_lf = np.zeros((n_points, 1))
d_velocity_lf = np.zeros((n_points, 1))

# --- Collect high fidelity points
for i in range(0, n_points):
    d_position_hf[i], d_velocity_hf[i] = mc.simulation(x_random[i, :])

# --- Collect low fidelity points  
for i in range(0, n_points):
    d_position_lf[i], d_velocity_lf[i] = mc.low_cost_simulation(x_random[i, :])

It is time to build the multi-fidelity model for both the position and the velocity.

As we did in the previous section we use the emulator to optimize the simulator. In this case we use the high fidelity output of the emulator.

### --- Optimize controller parameters 
obj_func = lambda x: mc.run_simulation(env, x)[0]
obj_func_emulator = lambda x: mc.run_emulation([position_model, velocity_model], x, car_initial_location)[0]
objective_multifidelity = GPyOpt.core.task.SingleObjective(obj_func)

And we optimize using Bayesian optimzation.

from GPyOpt.experiment_design.random_design import RandomDesign

model = GPyOpt.models.GPModel(optimize_restarts=5, verbose=False, exact_feval=True, ARD=True)
space= [{'name':'linear_1', 'type':'continuous', 'domain':(-1/1.2, +1)},
        {'name':'linear_2', 'type':'continuous', 'domain':(-1/0.07, +1/0.07)},
        {'name':'constant', 'type':'continuous', 'domain':(-1, +1)}]

design_space = GPyOpt.Design_space(space=space)
aquisition_optimizer = GPyOpt.optimization.AcquisitionOptimizer(design_space)

n_initial_points = 25
random_design = RandomDesign(design_space)
initial_design = random_design.get_samples(n_initial_points)
acquisition = GPyOpt.acquisitions.AcquisitionEI(model, design_space, optimizer=aquisition_optimizer)
evaluator = GPyOpt.core.evaluators.Sequential(acquisition)

bo_multifidelity = GPyOpt.methods.ModularBayesianOptimization(model, design_space, objective_multifidelity, acquisition, evaluator, initial_design)
bo_multifidelity.run_optimization(max_iter=50)

_, _, _, frames = mc.run_simulation(env, np.atleast_2d(bo_multifidelity.x_opt), render=True)
anim=mc.animate_frames(frames, 'Best controller with multi-fidelity emulator')

from IPython.core.display import HTML

HTML(anim.to_jshtml())

Best Controller with Multi-Fidelity Emulator

And problem solved! We see how the problem is also solved with 250 observations of the high fidelity simulator and 250 of the low fidelity simulator.

Emukit [edit]

Figure: The Emukit software is a set of software tools for emulation and surrogate modeling. https://amzn.github.io/emukit/

The aim is to provide a suite where different approaches to emulation are assimilated under one roof. The current version of Emukit includes multi-fidelity emulation for build surrogate models when data is obtained from multiple information sources that have different fidelity and/or cost; Bayesian optimisation for optimising physical experiments and tune parameters of machine learning algorithms or other computational simulations; experimental design and active learning: design the most informative experiments and perform active learning with machine learning models; sensitivity analysis: analyse the influence of inputs on the outputs of a given system; and Bayesian quadrature: efficiently compute the integrals of functions that are expensive to evaluate.

MXFusion: Modular Probabilistic Programming on MXNet [edit]

Figure: MXFusion is a probabilistic programming language targeted specifically at Gaussian process models and combining them with probaiblistic neural network. It is available through the MIT license and we welcome contributions throguh the Github repository https://github.com/amzn/MXFusion.

Work by Eric Meissner and Zhenwen Dai. Probabilistic programming. Available on Github

Figure: The MXFusion software.

MxFusion

Why another framework?

Key Requirements

Specialized inference methods + models, without requiring users to reimplement nor understand them every time. Leverage expert knowledge. Efficient inference, flexible framework. Existing frameworks either did one or the other: flexible, or efficient.

What does it look like?

Modelling

Inference

m = Model()
m.mu = Variable()
m.s = Variable(transformation=PositiveTransformation())
m.Y = Normal.define_variable(mean=m.mu, variance=m.s)

• Variable
• Distribution

• Function

• log_pdf

• draw_samples

• Variational Inference

• MCMC Sampling (soon) Built on MXNet Gluon (imperative code, not static graph)

infr = GradBasedInference(inference_algorithm=MAP(model=m, observed=[m.Y]))
infr.run(Y=data)

• Model + Inference together form building blocks.
  • Just doing modular modeling with universal inference doesn't really scale, need specialized inference methods for specialized modelling objects like non-parametrics.

Long term Aim

• Simulate/Emulate the components of the system.
  • Validate with real world using multifidelity.
  • Interpret system using e.g. sensitivity analysis.
• Perform end to end learning to optimize.
  • Maintain interpretability.

Stefanos Eleftheriadis, John Bronskill, Hugh Salimbeni, Rich Turner, Zhenwen Dai, Javier Gonzalez, Andreas Damianou, Mark Pullin, Eric Meissner.

• twitter: @lawrennd
• blog: http://inverseprobability.com

References