1. Extending Expectation Propagation for Graphical Models Yuan (Alan) Qi Joint work with Tom Minka

2. Motivation Graphical models are widely used in real- world applications, such as wireless communications and bioinformatics. Inference techniques on graphical models often sacrifice efficiency for accuracy or sacrifice accuracy for efficiency. Need a method that better balances the trade-off between accuracy and efficiency.

3. Motivation Computational Time Error Current Techniques What we want

4. Outline Background on expectation propagation (EP) Extending EP on Bayesian networks for dynamic systems –Poisson tracking –Signal detection for wireless communications Tree-structured EP on loopy graphs Conclusions and future work

5. Outline Background on expectation propagation (EP) Extending EP on Bayesian networks for dynamic systems –Poisson tracking –Signal detection for wireless communications Tree-structured EP on loopy graphs Conclusions and future work

6. Graphical Models Directed ( Bayesian networks) Undirected ( Markov networks) x1x1 x2x2 y1y1 y2y2 x1x1 x2x2 y1y1 y2y2

7. Inference on Graphical Models Bayesian inference techniques: –Belief propagation (BP): Kalman filtering /smoothing, forward-backward algorithm –Monte Carlo: Particle filter/smoothers, MCMC Loopy BP: typically efficient, but not accurate on general loopy graphs Monte Carlo: accurate, but often not efficient

8. Expectation Propagation in a Nutshell Approximate a probability distribution by simpler parametric terms: For directed graphs: For undirected graphs: Each approximation term lives in an exponential family (e.g. Gaussian)

9. EP in a Nutshell The approximate term minimizes the following KL divergence by moment matching: Where the leave-one-out approximation is

10. Limitations of Plain EP Can be difficult or expensive to analytically compute the needed moments in order to minimize the desired KL divergence. Can be expensive to compute and maintain a valid approximation distribution q(x), which is coherent under marginalization. –Tree-structured q(x):

11. Three Extensions 1. Instead of choosing the approximate term to minimize the following KL divergence: use other criteria. 2. Use numerical approximation to compute moments: Quadrature or Monte Carlo. 3. Allow the tree-structured q(x) to be non-coherent during the iterations. It only needs to be coherent in the end.

12. Efficiency vs. Accuracy Computational Time Error Extended EP ? Monte Carlo Loopy BP (Factorized EP)

13. Outline Background on expectation propagation (EP) Extending EP on Bayesian networks for dynamic systems –Poisson tracking –Signal detection for wireless communications Tree-structured EP on loopy graphs Conclusions and future work

14. Object Tracking Guess the position of an object given noisy observations Object

15. Bayesian Network (random walk) e.g. want distribution of x’s given y’s x1x1 x2x2 xTxT y1y1 y2y2 yTyT

16. Approximation Factorized and Gaussian in x

17. Message Interpretation = (forward msg)(observation msg)(backward msg) xtxt ytyt Forward Message Backward Message Observation Message

18. Extensions of EP Instead of matching moments, use any method for approximate filtering. –Examples: statistical linearization, unscented Kalman filter (UKF), mixture of Kalman filters Turn any deterministic filtering method into a smoothing method! All methods can be interpreted as finding linear/Gaussian approximations to original terms. Use quadrature or Monte Carlo for term approximations

19. Example: Poisson Tracking is an integer valued Poisson variate with mean

20. Poisson Tracking Model

21. Extended EP vs. Monte Carlo: Accuracy Variance Mean

22. Accuracy/Efficiency Tradeoff

23. Bayesian network for Wireless Signal Detection x1x1 x2x2 xTxT y1y1 y2y2 yTyT s1s1 s2s2 sTsT s i : Transmitted signals x i : Channel coefficients for digital wireless communications y i : Received noisy observations

24. Extended-EP Joint Signal Detection and Channel Estimation Turn mixture of Kalman filters into a smoothing method Smoothing over the last observations Observations before act as prior for the current estimation

25. Computational Complexity Expectation propagation O(nLd 2 ) Stochastic mixture of Kalman filters O(LMd 2 ) Rao-blackwellised particle smoothers O(LMNd 2 ) n: Number of EP iterations (Typically, 4 or 5) d: Dimension of the parameter vector L: Smooth window length M: Number of samples in filtering (Often larger than 500) N: Number of samples in smoothing (Larger than 50) EP is about 5,000 times faster than Rao-blackwellised particle smoothers.

26. Experimental Results EP outperforms particle smoothers in efficiency with comparable accuracy. (Chen, Wang, Liu 2000) Signal-Noise-Ratio

27. Bayesian Networks for Adaptive Decoding x1x1 x2x2 xTxT y1y1 y2y2 yTyT e1e1 e2e2 eTeT The information bits e t are coded by a convolutional error-correcting encoder.

28. EP Outperforms Viterbi Decoding Signal-Noise-Ratio

29. Outline Background on expectation propagation (EP) Extending EP on Bayesian networks for dynamic systems –Poisson tracking –Signal detection for wireless communications Tree-structured EP on loopy graphs Conclusions and future work

30. Inference on Loopy Graphs Problem: estimate marginal distributions of the variables indexed by the nodes in a loopy graph, e.g., p(x i ), i = 1,..., 16. X1X1 X2X2 X3X3 X4X4 X5X5 X6X6 X7X7 X8X8 X9X9 X 10 X 11 X 12 X 13 X 14 X 15 X 16

31. 4-node Loopy Graph Joint distribution is product of pairwise potentials for all edges: Want to approximate by a simpler distribution

32. BP vs. TreeEP BP TreeEP

33. Junction Tree Representation p(x) q(x) Junction tree

34. Two Kinds of Edges On-tree edges, e.g., (x 1,x 4 ): exactly incorporated into the junction tree Off-tree edges, e.g., (x 1,x 2 ): approximated by projecting them onto the tree structure

35. KL Minimization KL minimization moment matching Match single and pairwise marginals of Reduces to exact inference on single loops –Use cutset conditioning and

36. Matching Marginals on Graph (1) Incorporate edge ( x 3 x 4 ) x 3 x 4 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 (2) Incorporate edge ( x 6 x 7 ) x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 6 x 7 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6

37. Drawbacks of Global Propagation Update all the cliques even when only incorporating one off-tree edge –Computationally expensive Store each off-tree data message as a whole tree –Require large memory size

38. Solution: Local Propagation Allow q(x) be non-coherent during the iterations. It only needs to be coherent in the end. Exploit the junction tree representation: only locally propagate information within the minimal loop (subtree) that is directly connected to the off-tree edge. – Reduce computational complexity – Save memory

39. x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 3 x 4 x 1 x 2 x 1 x 3 x 1 x 4 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 5 x 7 x 1 x 2 x 1 x 3 x 1 x 4 x 3 x 5 x 3 x 6 x 3 x 5 x 6 x 7 x 5 x 7 x 3 x 5 x 3 x 6 (1) Incorporate edge( x 3 x 4 ) (3) Incorporate edge ( x 6 x 7 ) (2) Propagate evidence On this simple graph, local propagation runs roughly 2 times faster and uses 2 times less memory to store messages than plain EP

40. New Interpretation of TreeEP Marry EP with Junction algorithm Can perform efficiently over hypertrees and hypernodes

41. 4-node Graph TreeEP = the proposed method GBP = generalized belief propagation on triangles TreeVB = variational tree BP = loopy belief propagation = Factorized EP MF = mean-field

42. Fully-connected graphs Results are averaged over 10 graphs with randomly generated potentials TreeEP performs the same or better than all other methods in both accuracy and efficiency!

43. 8x8 grids, 10 trials MethodFLOPSError Exact30,0000 TreeEP300,0000.149 BP/double-loop15,500,0000.358 GBP17,500,0000.003

44. TreeEP versus BP and GBP TreeEP is always more accurate than BP and is often faster TreeEP is much more efficient than GBP and more accurate on some problems TreeEP converges more often than BP and GBP

45. Outline Background on expectation propagation (EP) Extending EP on Bayesian networks for dynamic systems –Poisson tracking –Signal detection for wireless communications Tree-structured EP on loopy graphs Conclusions and future work

46. Conclusions Extend EP on graphical models: –Instead of minimizing KL divergence, use other sensible criteria to generate messages. Effectively turn any deterministic filtering method into a smoothing method. –Use quadrature to approximate messages. –Local propagation to save the computation and memory in tree structured EP.

47. Conclusions Extended EP algorithms outperform state-of-art inference methods on graphical models in the trade- off between accuracy and efficiency Computational Time Error Extended EP State-of-art Techniques

48. Future Work More extensions of EP: –How to choose a sensible approximation family (e.g. which tree structure) –More flexible approximation: mixture of EP? –Error bound? –Bayesian conditional random fields –EP for optimization (generalize max-product) More real-world applications, e.g., classification of gene expression data.

49. Classifying Colon Cancer Data by Predictive Automatic Relevance Determination The task: distinguish normal and cancer samples The dataset: 22 normal and 40 cancer samples with 2000 features per sample. The dataset was randomly split 100 times into 50 training and 12 testing samples. SVM results from Li et al. 2002

50. End Contact information: yuanqi@media.mit.edu