# Graphical model / machine learning decoder ring

January 18, 2024 — February 29, 2024

algebra
graphical models
hidden variables
hierarchical models
how do science
machine learning
networks
neural nets
probability
statistics

I’m thinking something through for myself. Details are absent, right now. Twin to Causality+ML, perhaps.

Various ways of partitioning a vector of observed and unobserved variates, and their relationships to graphical models and the ML distinction of supervised/unsupervised learning, and the generalisations we might want to make these go.

## 1 What is supervised learning?

We look to Bayesian inference to solve problems with the following structure: I would like to infer the probability of some noisy label $$Y_*$$ given some predictor $$X_*$$. Since this is data-driven learning I will moreover suppose that I want to work out how to do this from a dataset of labelled examples, $$\mathcal{D} = \{(x_i, y_i)\}_{i=1}^N$$. So in fact, I would like to know $$p(y_*|x_*,\mathcal{D})$$.

We are using lazy Bayesian density notation where $$p(z)$$ is the density of some RV $$Z$$. I feel ambivalent about this notation, but it does get the job done. We can throw out the densities later and go to nice clean unambiguous measures, but this helps exposition.

How do I compute this equation-in-densities? Let us suppose that by some miracle it happens to be true that we know the generating process $$Y= G(X, W, \xi)$$. where $$\xi$$ is some unobserved noise, and there is a finite vector of of parameters $$W$$.

I hope to find good values for $$W$$ such that $$p(y_*|x_*,W,\mathcal{D})$$ is a good model for the data, in that if we take our model out into the world and show it lots of different values for $$x_*=X_*$$ we will observe that the implied $$p(y_*|X_*=x_*,W,\mathcal{D})$$, pumped through $$G$$, describes the distribution of $$Y_*|X_*=x_*$$.

How do we find $$W$$? There is the graphical model formalism which (details to come, trust me) tells us how to infer stuff about $$W$$ from $$\mathcal{D}$$: Suppose $$N=3$$. Then the following graphical model describes our task:

The circled $$W$$ node is what we are trying to get “correct”. By looking at those $$X_i,Y_i$$ pairs we refine our estimate of $$W$$. I am leaving out the details of what “refining” the estimate means; we will come back to that.

OK, we talked about refining $$W$$. What does that look like in practice? In the Bayes setting, we condition a prior $$p(w)$$ on the observed data $$\mathcal{D}$$, to obtain the posterior distribution $$p(w|\mathcal{D})=p(w|X_1, X_2, X_3, Y_1, Y_2, Y_3)$$. I am pretty sure we can alternatively do this as a frequentist thing, but TBH I am just not smart enough; the phrasing gets very weird, and we need to use all kinds of circumlocutions so it is not very clear, at least for my modest-sized brain what the hell is going on.

According to Bayes’ rule, we can write

$p(w \mid x_1, x_2, x_3, y_1, y_2, y_3) = \frac{p(y_1, y_2, y_3 \mid x_1, x_2, x_3, w) \, p(w)}{p(y_1, y_2, y_3 \mid x_1, x_2, x_3)}$

The numerator can be further factorized due to the conditional independencies implied by the graphical model:

$p(y_1, y_2, y_3 \mid x_1, x_2, x_3, w) \, p(w) = p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w)$

The denominator, which is the marginal likelihood, can be obtained by integrating out $$W$$:

$p(y_1, y_2, y_3 \mid x_1, x_2, x_3) = \int p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w) \, dw$

Putting it all together, the posterior distribution of $$W$$ is:

\begin{aligned} p(w \mid x_1, x_2, x_3, y_1, y_2, y_3) &= \frac{p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w)}{\int p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w) \, dw}\\ &= \frac{p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w)}{\int p(y_1 \mid x_1, w) \, p(y_2 \mid x_2, w) \, p(y_3 \mid x_3, w) \, p(w) \, dw} \end{aligned}

This equation represents the posterior distribution of $$W$$ in terms of the densities of the observed and latent variables, conditioned on the observed data.

This is misleading, since I said before we wanted a $$W$$ that gave us good predictions for $$Y_*|X_*$$. Maybe $$W$$ is just a nuisance parameter and we don’t really care about it as long as we get good predictions for $$Y_*$$. Maybe we actually want to solve a problem like this, and that $$W$$ is a nuisance variable:

In ML, that is usually what we are doing; $$W$$ is just some parameters rather than being something with intrinsic, semi-universal meaning, like the speed of light or the boiling point of lead in a vacuum. Our true target is to get this $$p(y_*|x_*,\mathcal{D})$$, which we give the special name, posterior predictive distribution.

Let us replay all that conditioning mathematics, but make the target $$p(y_*|x_*,\mathcal{D})$$, the posterior $$Y_*$$ given $$X_*$$, and the observed data $$X_1, Y_1, \ldots,$$, i.e.

$p(y_* \mid x_*, w, x_1, y_1, \ldots, )$

Since information about $$Y_i$$ and $$X_i$$ about $$Y_*$$ is mediated wholly through $$W$$, we might like to think about that calculation in terms of a posterior $$W$$,

$p(y_* \mid x_*, (W|\mathcal{D}))=:p(y_* \mid x_*, x_1, y_1, \ldots)$

Hold on to that thought, because the idea of dealing with “conditioned” variables turns out to take us somewhere interesting.

If we want to integrate out $$W$$ to get only the marginal posterior predictive distribution, we do this:

$p(y_* \mid x_*, x_1, y_1, \ldots) = \int p(y_* \mid x_*, w) \, p(w \mid x_1, y_1, \ldots) \, dw$

Where $$p(w \mid x_1, y_1, \ldots, )$$ is the posterior distribution of $$W$$ given the observed data, which can be computed using Bayes’ rule as we did above.

I’m getting tired of writing out the indexed $$X_i,Y_i$$ pairs, so let’s just write $$\mathcal{D}$$ for the observed data, and $$X_*,Y_*$$ for the unobserved data, and use plate notation to write a graph that automatically indexes over these variables:

OK, now in a Bayes setting we can talk about the marginal of interest. The red line denotes the marginal of interest.

## 2 What is unsupervised learning?

As seen in unconditional generative modeling. Now we don’t assume that we will observe a special $$X_*$$ and find $$Y_*$$, but rather we want to know something about the distribution of both jointly

In fact, we are treating $$X$$ and $$Y$$ symmetrically now, so we might as well concatenate them both into $$X$$:

## 3 What are inverse problems?

I observe an output. What is my “posterior retrodictive”?

## 4 Other cool problems?

Hierarchical models.

TBD

## 7 Variational belief propagation (speed run)

We say we can solve one of these problems if we can propagate information through the nodes of the graph and find the target marginal distribution, in some sense. This might be the target marginal in the sense that we can sample from it (a generative model) or in the sense that we can compute some density or expectation (a discriminative model). For now, we are agnostic about that.

Even with that relaxed notion of “marginalisation” we do not necessarily have the power to solve these. Propagating information through these nodes is not generally tractable. Some special cases are tractable, but actually, let us leap into the void and just seek approximate solutions.

Normally at this point we laboriously introduce many different inference methods, junction trees and identifiability criteria and so on. Then we say “OH NO BUT THAT DOES NOT ACTUALLY WORK ON MY PROBLEM!!!1!”.

There are usually two classes of difficulties with classical algorithms:

1. Our favoured algorithm is only exact on tree graphs, but we have a graph with cycles.
2. Our favoured algorithm only works on nodes whose joint distributions at each node factor into exponential families at each edge.

Few interesting problems satisfy both of these conditions.

So let us speed run through the hand-wringing bit where we worry that we are only solving integrals approximately, and just go to some approximate inference methods that do pretty well in practical problems; we can sweat the details later.

We will introduce one, concrete inference method, variational belief propagation, which will crystallise understanding of the general problem of approximate inference, and work outwards from there.

There is some necessary machinery

## 8 Factor graphs

A natural setting for approximate inference is to devise variational algorithms over factor graphs which conveniently encode some useful approximate marginalisation algorithms.

The way that these work is we rewrite every conditional probability in the directed graph as a factor:

$p(x_1|x_2) \to f_{12}(x_1, x_2)$

Now, we introduce a new set of nodes, the factors, which are the $$f_{ij}$$, and we connect them to the implicated variable nodes.

Why? Everyone asks me why. My answer for now is that it comes out easier that way so suck it up.

A nicer answer might be: the various belief propagation rules come out nice this way, if we consider the variables and their relations independently.

Here is the $$W$$ inference problem as a factor graph:

Here is that inverse problem from earlier.

We can spend a lot of time circumlocuting about when we can actually update over such a graph. The answer is complicated, but we are neural network people, so we’ll just YOLO it see what happens.

## 9 YOLO belief propagation

Call the entire variable vector $$X$$. Let us partition it and work out what we can do with those partitions. Suppose we have some factorisation of the density over it, so that the joint over all variates is

$p(X)=\frac{1}{Z} \prod_a f_a\left(X_a\right)=\frac{1}{Z} \prod_a e^{-E\left(X_a\right)} .$

We will start out doing this heuristically, to build intuitions.

👷

### 9.2 BP as variational approximation

Our goal in variational inference is to find an equation in beliefs, which are approximate marginals, and hope that we can calculate them in such a way that they end up good approximations for true marginals.

We want to represent every quantity of interest in terms of marginals at each node, and then solve for those marginals. If we can find some fixed point iteration that is local in some sense, and promises convergence to something useful, then we can declare victory.

Great question.

## 11 To touch upon

• Dense non-causal connections
• re-factorisation
• inference estimate
• (approximate) independence and identifiability are simpler as variational approximation problems

## 12 References

Akbayrak, Semih, Ivan Bocharov, and Bert de Vries. 2021. Entropy 23 (7): 815.
Attias, Hagai. 1999. In Proceedings of the Fifteenth Conference on Uncertainty in Artificial Intelligence, 21–30. UAI’99. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.
Blei, David M., Alp Kucukelbir, and Jon D. McAuliffe. 2017. Journal of the American Statistical Association 112 (518): 859–77.
Cox, Marco, Thijs van de Laar, and Bert de Vries. 2019. International Journal of Approximate Reasoning 104 (January): 185–204.
Dauwels, Justin. 2007. In 2007 IEEE International Symposium on Information Theory, 2546–50. Nice: IEEE.
Jordan, Michael I. 2004. Statistical Science 19 (1): 140–55.
Jordan, Michael I., Zoubin Ghahramani, Tommi S. Jaakkola, and Lawrence K. Saul. 1999. Machine Learning 37 (2): 183–233.
Kschischang, Frank R, Brendan J Frey, and Hans-Andrea Loeliger. 2001. IEEE Transactions on Information Theory 47 (2): 498–519.
Kuck, Jonathan, Shuvam Chakraborty, Hao Tang, Rachel Luo, Jiaming Song, Ashish Sabharwal, and Stefano Ermon. 2020. arXiv:2007.00295 [Cs, Stat], July.
Li, Yingzhen, and Richard E Turner. 2016. In Advances in Neural Information Processing Systems, 29:1081–89. Red Hook, NY, USA: Curran Associates, Inc.
Malioutov, Dmitry M., Jason K. Johnson, and Alan S. Willsky. 2006. Journal of Machine Learning Research 7 (October): 2031–64.
Minka, Thomas P. 2001. In Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence, 362–69. UAI’01. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.
Naesseth, Christian Andersson, Fredrik Lindsten, and Thomas B Schön. 2014. In Advances in Neural Information Processing Systems. Vol. 27. Curran Associates, Inc.
Nguyen, Trung V., and Edwin V. Bonilla. 2014. In Proceedings of the 27th International Conference on Neural Information Processing Systems - Volume 1, 1404–12. NIPS’14. Cambridge, MA, USA: MIT Press.
Ortiz, Joseph, Talfan Evans, and Andrew J. Davison. 2021. arXiv:2107.02308 [Cs], July.
Pearl, Judea. 2009. Causality: Models, Reasoning and Inference. Cambridge University Press.
Roychowdhury, Anirban, and Brian Kulis. 2015. In Artificial Intelligence and Statistics, 800–808. PMLR.
Satorras, Victor Garcia, and Max Welling. 2021. In. arXiv.
Wainwright, Martin J., and Michael I. Jordan. 2008. Graphical Models, Exponential Families, and Variational Inference. Vol. 1. Foundations and Trends® in Machine Learning. Now Publishers.
Wand, M. P. 2017. Journal of the American Statistical Association 112 (517): 137–68.
Welling, Max, Tom P Minka, and Yee Whye Teh. 2012. In Uncertainty in Artificial Intelligence.
Wiegerinck, Wim, and Tom Heskes. 2002. In Advances in Neural Information Processing Systems. Vol. 15. Cambridge, MA, USA: MIT Press.
Wingate, David, and Theophane Weber. 2013. arXiv:1301.1299 [Cs, Stat], January.
Winn, John M., and Christopher M. Bishop. 2005. In Journal of Machine Learning Research, 661–94.
Xing, Eric P., Michael I. Jordan, and Stuart Russell. 2003. In Proceedings of the Nineteenth Conference on Uncertainty in Artificial Intelligence, 583–91. UAI’03. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.
Yedidia, Jonathan S., William T. Freeman, and Yair Weiss. 2000. “Generalized Belief Propagation.” In Proceedings of the 13th International Conference on Neural Information Processing Systems, 668–74. NIPS’00. Cambridge, MA, USA: MIT Press.
Yedidia, J.S., W.T. Freeman, and Y. Weiss. 2003. In Exploring Artificial Intelligence in the New Millennium, edited by G. Lakemeyer and B. Nebel, 239–36. Morgan Kaufmann Publishers.
———. 2005. IEEE Transactions on Information Theory 51 (7): 2282–312.
Zhou, Hai-Jun. 2017. 15.