Sequential Monte Carlo and Particle Filtering. Frank Wood Gatsby, November 2007
|
|
- Silas Summers
- 6 years ago
- Views:
Transcription
1 Sequential Monte Carlo and Particle Filtering Frank Wood Gatsby, November 2007
2 Importance Sampling Recall: Let s say that we want to compute some expectation (integral) E p [f] = p(x)f(x)dx and we remember from Monte Carlo integration theory that with samples from p() we can approximate this integral thusly E p [f] 1 L L f(x l ) l=1
3 What if p() is hard to sample from? One solution: use importance sampling Choose another easier to sample distribution q() that is similar to p() and do the following: E p [f] = p(x)f(x)dx = 1 L p(x) q(x) f(x)q(x)dx L p(x l ) q(x l ) f(x l) l=1 x l q( )
4 I.S. with distributions known only to a proportionality Importance sampling using distributions only known up to a proportionality is easy and common, algebra yields E p [f] Z q Z p 1 L L l=1 L l=1 w l f(x l ) p(x l ) q(x l ) f(x l) r l = p(x l) q(x l ) w l = r l L l=1 r l
5 A model requiring sampling techniques Non-linear non-gaussian first order Markov model x t 1 \\ x t x t+1 Hidden and of interest y t 1 y t y t+1 p(x 1:i,y 1:i ) = N i=1 p(y i x i )p(x i x i 1 )
6 Filtering distribution hard to obtain Often the filtering distribution is of interest p(x i y 1:i 1 )... p(x 1:i,y 1:i )d x1... d xi 1 It may not be possible to compute these integrals analytically, be easy to sample from this directly, nor even to design a good proposal distribution for importance sampling.
7 A solution: sequential Monte Carlo Sample from sequence of distributions that converge to the distribution of interest This is a very general technique that can be applied to a very large number of models and in a wide variety of settings. Today: particle filtering for a first order Markov model
8 Concrete example: target tracking A ballistic projectile has been launched in our direction. We have orders to intercept the projectile with a missile and thus need to infer the projectiles current position given noisy measurements.
9 Problem Schematic (x t, y t ) r t θ t (0,0)
10 Probabilistic approach Treat true trajectory as a sequence of latent random variables Specify a model and do inference to recover the position of the projectile at time t
11 First order Markov model x t 1 x t x t+1 Hidden true trajectory y t 1 y t y t+1 Noisy measurements p(x t+1 x t ) = a( ; θ a, x t ) p(y t x t ) = h( ; θ h, x t )
12 Posterior expectations We want filtering distribution samples p(x i y 1:i ) p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 ) Write this down! so that we can compute expectations E p [f] = f(x i )p(x i y 1:i )dx i
13 Importance sampling p(x i ) p(y i x i )p(x i y 1:i 1 ) q(x i ) = p(x i y 1:i 1 ) Distribution from which we want samples Proposal distribution By identity the posterior predictive distribution can be written as q(x i ) = p(x i y 1:i 1 ) = p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1
14 Basis of sequential recursion If we start with samples from {w i 1 l, x i 1 l } L l=1 p(x i 1 y 1:i 1 ) then we can write the proposal distribution as a finite mixture model L q(x i ) = p(x i y 1:i 1 ) w i 1 l p(x i x i 1 l ) l=1 and draw samples accordingly {ŵ i m, ˆx i m} M m=1 q( )
15 Samples from the proposal distribution We now have samples from the proposal {ŵ i m, ˆx i m} q(x i ) And if we recall p(x i ) = p(y i x i )p(x i y 1:i 1 ) q(x i ) = p(x i y 1:i 1 ) Distribution from which we want samples Proposal distribution
16 Updating the weights completes importance sampling ˆr i m = p(ˆxi m ) q(ˆx i m ) = p(y i ˆx i m)ŵ i m We are left with M weighted samples from the posterior up to observation i w i m = Mˆr i m m=1ˆri m {w i m, x i m} M m=1 p(x i y 1:i )
17 Intuition Particle filter name comes from physical interpretation of samples
18 Start with samples representing the hidden state {w i 1 l, x i 1 l } L l=1 p(x i 1 y 1:i 1 ) (0,0)
19 Evolve them according to the state model ˆx i m p( x i 1 l ) ŵ i m w i 1 m (0,0)
20 Re-weight them by the likelihood w i m ŵ i mp(y i ˆx i m) (0,0)
21 Results in samples one step forward {w i l, xi l }L l=1 {wi m, x i m} M m=1 (0,0)
22 SIS Particle Filter The Sequential Importance Sampler (SIS) particle filter multiplicatively updates weights at every iteration and thus often most weights get very small Computationally this makes little sense as eventually low-weighted particles do not contribute to any expectations. A measure of particle degeneracy is effective sample size ˆN eff = 1 L l=1 (wi l )2 When this quantity is small, particle degeneracy is severe.
23 Solutions Sequential Importance Re-sampling (SIR) particle filter avoids many of the problems associated with SIS pf ing by re-sampling the posterior particle set to increase the effective sample size. Choosing the best possible importance density is also important because it minimizes the variance of the weights which maximizes ˆNeff
24 Other tricks to improve pf ing Integrate out everything you can Replicate each particle some number of times In discrete systems, enumerate instead of sample Use fancy re-sampling schemes like stratified sampling, etc.
25 Initial particles {w i 1 l, x i 1 l } L l=1 p(x i 1 y 1:i 1 ) (0,0)
26 Particle evolution step {ŵ i m, ˆx i m} M m=1 p(x i y 1:i 1 ) (0,0)
27 Weighting step {w i m, x i m} M m=1 p(x i y 1:i ) (0,0)
28 Resampling step {w i l, xi l }L l=1 p(x i y 1:i ) (0,0)
29 Wrap-up: Pros vs. Cons Pros: Sometimes it is easier to build a good particle filter sampler than an MCMC sampler No need to specify a convergence measure Cons: Really filtering not smoothing Issues Computational trade-off with MCMC
30 Thank You
31
32
33 Tricks and Variants Reduce the dimensionality of the integrand through analytic integration Rao-Blackwellization Reduce the variance of the Monte Carlo estimator through Maintaining a weighted particle set Stratified sampling Over-sampling Optimal re-sampling
34 Particle filtering Consists of two basic elements: Monte Carlo integration L p(x) w l δ xl lim L L l=1 w l f(x l ) = Importance sampling l=1 f(x)p(x)dx
35 PF ing: Forming the posterior predictive {w i 1 l, x i 1 l } L l=1 p(x i 1 y 1:i 1 ) p(x i y 1:i 1 ) = Posterior up to observation i 1 p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 L l=1 w i 1 l p(x i x i 1 l ) The proposal distribution for importance sampling of the posterior up to observation i is this approximate posterior predictive distribution
36 Sampling the posterior predictive Generating samples from the posterior predictive distribution is the first place where we can introduce variance reduction techniques {ŵ i m, ˆxi m }M m=1 p(x i y 1:i 1 ), p(x i y 1:i 1 ) For instance sample from each mixture component several twice such that M, the number of samples drawn, is two times L, the number of densities in the mixture model, and assign weights ŵ i m = wi 1 l 2 L l=1 w i 1 l p(x i x i 1 l )
37 Not the best Most efficient Monte Carlo estimator of a function Γ(x) From survey sampling theory: Neyman allocation Number drawn from each mixture density is proportional to the weight of the mixture density times the std. dev. of the function Γ over the mixture density Take home: smarter sampling possible [Cochran 1963]
38 Over-sampling from the posterior predictive distribution {ŵ i m, ˆx i m} M m=1 p(x i y 1:i 1 ) (0,0)
39 Importance sampling the posterior Recall that we want samples from p(x i y 1:i ) p(y i x i ) p(y i x i )p(x i y 1:i 1 ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 and make the following importance sampling identifications p(x i ) = p(y i x i )p(x i y 1:i 1 ) q(x i ) = p(x i y 1:i 1 ) L Distribution from which we want to sample Proposal distribution l=1 w i 1 l p(x i x i 1 l )
40 Sequential importance sampling Weighted posterior samples arise as {ŵ i m, ˆx i m} q( ) Normalization of the weights takes place as before wm i ˆr = i l L l=1ˆri l ˆr i m = p(ˆxi m ) q(ˆx i m ) = p(y i ˆx i m)ŵ i m We are left with M weighted samples from the posterior up to observation i {w i m, x i m} M m=1 p(x i y 1:i )
41 An alternative view {ŵ i m, ˆx i m} L l=1 w i 1 l p(x i x i 1 l ) p(x i y 1:i 1 ) M m=1 ŵi mδˆx i m p(x i y 1:i ) M m=1 p(y i ˆx i m)ŵ i mδˆx i m
42 Importance sampling from the posterior distribution {w i m, x i m} M m=1 p(x i y 1:i ) (0,0)
43 Sequential importance re-sampling Down-sample L particles and weights from the collection of M particles and weights {w i l, xi l }L l=1 {wi m, x i m} M m=1 this can be done via multinomial sampling or in a way that provably minimizes estimator variance [Fearnhead 04]
44 Down-sampling the particle set {w i l, xi l }L l=1 p(x i y 1:i ) (0,0)
45 Recap Starting with (weighted) samples from the posterior up to observation i-1 Monte Carlo integration was used to form a mixture model representation of the posterior predictive distribution The posterior predictive distribution was used as a proposal distribution for importance sampling of the posterior up to observation i M > L samples were drawn and re-weighted according to the likelihood (the importance weight), then the collection of particles was down-sampled to L weighted samples
46 LSSM Not alone Various other models are amenable to sequential inference, Dirichlet process mixture modelling is another example, dynamic Bayes nets are another
47 Rao-Blackwellization In models where parameters can be analytically marginalized out, or the particle state space can otherwise be collapsed, the efficiency of the particle filter can be improved by doing so
48 Stratified Sampling Sampling from a mixture density using the algorithm on the right produces a more efficient Monte Carlo estimator {w n, x n } K n=1 k π k f k ( ) for n=1:k choose k according to π k sample x n from f k set w n equal to 1/N for n=1:k choose f n sample x n from f n set w n equal to π n
49 Intuition: weighted particle set What is the difference between these two discrete distributions over the set {a,b,c}? (a), (a), (b), (b), (c) (.4, a), (.4, b), (.2, c) Weighted particle representations are equally or more efficient for the same number of particles
50 Optimal particle re-weighting Next step: when down-sampling, pass all particles above a threshold c through without modifying their weights where c is the unique solution to M m=1 min{cwi m, 1} = L Resample all particles with weights below c using stratified sampling and give the selected particles weight 1/ c Fearnhead 2004
51 Result is provably optimal In the down-sampling step {w i l, xi l }L l=1 {wi m, x i m} M m=1 Imagine instead a sparse set of weights of which some are zero { w i l, xi l }M l=1 {wi m, x i m} M m=1 Then this down-sampling algorithm is optimal w.r.t. M m=1 E w[( w i m w i m) 2 ] Fearnhead 2004
52 Problem Details Time and position are given in seconds and meters respectively Initial launch velocity and position are both unknown The maximum muzzle velocity of the projectile is 1000m/s The measurement error in the Cartesian coordinate system is N(0,10000) and N(0,500) for x and y position respectively The measurement error in the polar coordinate system is N(0,.001) for θ and Gamma(1,100) for r The kill radius of the projectile is 100m
53 Data and Support Code
54 Laws of Motion In case you ve forgotten: r = (v 0 cos(α))ti + ((v 0 sin(α)t 1 2 gt2 )j where v 0 is the initial speed and α is the initial angle
55 Good luck!
56 Monte Carlo Integration Compute integrals for which analytical solutions are unknown f(x)p(x)dx p(x) L w l δ xl l=1
57 Monte Carlo Integration Integral approximated as the weighted sum of function evaluations at L points f(x)p(x)dx p(x) L w l δ xl l=1 = L l=1 f(x) L l=1 w l f(x l ) w l δ xl dx
58 Sampling To use MC integration one must be able to sample from p(x) lim L {w l, x l } L l=1 p( ) L w l δ xl p( ) l=1
59 Theory (Convergence) Quality of the approximation independent of the dimensionality of the integrand Convergence of integral result to the truth is O(1/n 1/2 ) from L.L.N. s. Error is independent of dimensionality of x
60 Bayesian Modelling Formal framework for the expression of modelling assumptions Two common definitions: using Bayes rule marginalizing over models Posterior Prior p(θ x) = p(x θ)p(θ) p(x) = p(x θ)p(θ) p(x θ)p(θ)dθ Likelihood Evidence
61 Posterior Estimation Often the distribution over latent random variables (parameters) is of interest Sometimes this is easy (conjugacy) Usually it is hard because computing the evidence is intractable
62 Conjugacy Example θ Beta(α, β) x θ Binomial(N, θ) p(θ) = p(x θ) = Γ(α + β) Γ(α)Γ(β) θα 1 (1 θ) β 1 ( ) N θ x (1 θ) N x x x successes in N trials, θ probability of success
63 Conjugacy Continued p(θ x) = = = p(x θ)p(θ) p(x θ)p(θ)dθ 1 Z(x) p(x θ)p(θ) 1 Z(x) θα 1+x (1 θ) β 1+N x θ x Beta(α + x, β + N x) ( Γ(α + β + N) Z(x) = Γ(α + x)γ(β + N x) ) 1
64 Non-Conjugate Models Easy to come up with examples σ 2 N(0, α) x σ 2 N(0, σ 2 )
65 Posterior Inference Posterior averages are frequently important in problem domains posterior predictive distribution p(x i+1 x 1:i ) = p(x i+1 θ,x 1:i )p(θ x 1:i )dθ evidence (as seen) for model comparison, etc.
66 Relating the Posterior to the Posterior Predictive p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
67 Relating the Posterior to the Posterior Predictive p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
68 Relating the Posterior to the Posterior Predictive p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
69 Relating the Posterior to the Posterior Predictive p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
70 Importance sampling E x [f(x)] = p(x)f(x)dx = p(x) q(x) f(x)q(x)dx 1 L p(x l ) L q(x l ) f(x l) Original distribution: hard to sample from, easy to evaluate l=1 Proposal distribution: easy to sample from x l q( ) Importance weights r l = p(x l) q(x l )
71 Importance sampling un-normalized distributions p(x) = p(x) Z p q(x) = q(x) Z q Un-normalized distribution to sample from, still hard to sample from and easy to evaluate x l q( ) Un-normalized proposal distribution: still easy to sample from New term: ratio of normalizing constants E x [f(x)] 1 L L l=1 Z q Z p 1 L p(x l ) q(x l ) f(x l) L l=1 p(x l ) q(x l ) f(x l)
72 Normalizing the importance weights Un-normalized importance weights Takes a little algebra r l = p(x l) q(x l ) Z q L Z p L l=1 r l w l = r l L l=1 r l Normalized importance weights E x [f(x)] Z q Z p 1 L L l=1 L l=1 w l f(x l ) p(x l ) q(x l ) f(x l)
73 Linear State Space Model (LSSM) Discrete time First-order Markov chain x t+1 = ax t + ǫ, ǫ N(µ ǫ, σǫ) 2 y t = bx t + η, η N(µ η, ση) 2 x t 1 x t x t+1 y t 1 y t y t+1
74 Inferring the distributions of interest Many methods exist to infer these distributions Markov Chain Monte Carlo (MCMC) Variational inference Belief propagation etc. In this setting sequential inference is possible because of characteristics of the model structure and preferable due to the problem requirements
75 Exploiting LSSM model structure p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
76 Particle filtering p(x i y 1:i ) p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1
77 Exploiting Markov structure p(x i y 1:i ) = Use p(xbayes i,x 1:i 1 rule yand 1:i the )dxconditional 1:i 1 independence structure dictated by the first order Markov hidden variable model p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 x t 1 x t x t+1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x y t 1 i ) p(x i y x t i 1 )p(x y t+1 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
78 for sequential inference p(x i y 1:i ) = p(x i,x 1:i 1 y 1:i )dx 1:i 1 p(y i x i,x 1:i 1,y 1:i 1 )p(x i,x 1:i 1,y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x 1:i 1,y 1:i )p(x 1:i 1 y 1:i )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x 1:i 1 y 1:i 1 )dx 1:i 1 p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 )
79 An alternative view {ŵ i m, ˆx i m} L l=1 w i 1 l p(x i x i 1 l ) p(x i y 1:i 1 ) M m=1 ŵi mδˆx i m p(x i y 1:i ) M m=1 p(y i ˆx i m)ŵ i mδˆx i m
80 Sequential importance sampling inference Start with a discrete representation of the posterior up to observation i-1 Use Monte Carlo integration to represent the posterior predictive distribution as a finite mixture model Use importance sampling with the posterior predictive distribution as the proposal distribution to sample the posterior distribution up to observation i
81 Posterior predictive distribution What? State model Likelihood p(x i y 1:i ) p(y i x i ) p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 p(y i x i )p(x i y 1:i 1 ) Start with a discrete representation of this distribution
82 Monte Carlo integration General setup p(x) L l=1 w l δ xl lim L L l=1 w l f(x l ) = f(x)p(x)dx As applied in this stage of the particle filter p(x i y 1:i 1 ) = p(x i x i 1 )p(x i 1 y 1:i 1 )dx i 1 L l=1 w i 1 l p(x i x i 1 l ) Samples from Finite mixture model
Particle Filtering a brief introductory tutorial. Frank Wood Gatsby, August 2007
Particle Filtering a brief introductory tutorial Frank Wood Gatsby, August 2007 Problem: Target Tracking A ballistic projectile has been launched in our direction and may or may not land near enough to
More informationAnswers and expectations
Answers and expectations For a function f(x) and distribution P(x), the expectation of f with respect to P is The expectation is the average of f, when x is drawn from the probability distribution P E
More informationL09. PARTICLE FILTERING. NA568 Mobile Robotics: Methods & Algorithms
L09. PARTICLE FILTERING NA568 Mobile Robotics: Methods & Algorithms Particle Filters Different approach to state estimation Instead of parametric description of state (and uncertainty), use a set of state
More informationDensity Estimation. Seungjin Choi
Density Estimation Seungjin Choi Department of Computer Science and Engineering Pohang University of Science and Technology 77 Cheongam-ro, Nam-gu, Pohang 37673, Korea seungjin@postech.ac.kr http://mlg.postech.ac.kr/
More informationVariational Scoring of Graphical Model Structures
Variational Scoring of Graphical Model Structures Matthew J. Beal Work with Zoubin Ghahramani & Carl Rasmussen, Toronto. 15th September 2003 Overview Bayesian model selection Approximations using Variational
More informationMCMC and Gibbs Sampling. Kayhan Batmanghelich
MCMC and Gibbs Sampling Kayhan Batmanghelich 1 Approaches to inference l Exact inference algorithms l l l The elimination algorithm Message-passing algorithm (sum-product, belief propagation) The junction
More informationIntroduction to Bayesian inference
Introduction to Bayesian inference Thomas Alexander Brouwer University of Cambridge tab43@cam.ac.uk 17 November 2015 Probabilistic models Describe how data was generated using probability distributions
More informationLecture 2: From Linear Regression to Kalman Filter and Beyond
Lecture 2: From Linear Regression to Kalman Filter and Beyond January 18, 2017 Contents 1 Batch and Recursive Estimation 2 Towards Bayesian Filtering 3 Kalman Filter and Bayesian Filtering and Smoothing
More informationIntroduction to Probabilistic Machine Learning
Introduction to Probabilistic Machine Learning Piyush Rai Dept. of CSE, IIT Kanpur (Mini-course 1) Nov 03, 2015 Piyush Rai (IIT Kanpur) Introduction to Probabilistic Machine Learning 1 Machine Learning
More informationLearning Bayesian network : Given structure and completely observed data
Learning Bayesian network : Given structure and completely observed data Probabilistic Graphical Models Sharif University of Technology Spring 2017 Soleymani Learning problem Target: true distribution
More informationCPSC 540: Machine Learning
CPSC 540: Machine Learning MCMC and Non-Parametric Bayes Mark Schmidt University of British Columbia Winter 2016 Admin I went through project proposals: Some of you got a message on Piazza. No news is
More informationIntroduction to Systems Analysis and Decision Making Prepared by: Jakub Tomczak
Introduction to Systems Analysis and Decision Making Prepared by: Jakub Tomczak 1 Introduction. Random variables During the course we are interested in reasoning about considered phenomenon. In other words,
More informationUnsupervised Learning
Unsupervised Learning Bayesian Model Comparison Zoubin Ghahramani zoubin@gatsby.ucl.ac.uk Gatsby Computational Neuroscience Unit, and MSc in Intelligent Systems, Dept Computer Science University College
More informationBayesian Methods for Machine Learning
Bayesian Methods for Machine Learning CS 584: Big Data Analytics Material adapted from Radford Neal s tutorial (http://ftp.cs.utoronto.ca/pub/radford/bayes-tut.pdf), Zoubin Ghahramni (http://hunch.net/~coms-4771/zoubin_ghahramani_bayesian_learning.pdf),
More informationDavid Giles Bayesian Econometrics
David Giles Bayesian Econometrics 1. General Background 2. Constructing Prior Distributions 3. Properties of Bayes Estimators and Tests 4. Bayesian Analysis of the Multiple Regression Model 5. Bayesian
More informationPart 1: Expectation Propagation
Chalmers Machine Learning Summer School Approximate message passing and biomedicine Part 1: Expectation Propagation Tom Heskes Machine Learning Group, Institute for Computing and Information Sciences Radboud
More informationApproximate Bayesian inference
Approximate Bayesian inference Variational and Monte Carlo methods Christian A. Naesseth 1 Exchange rate data 0 20 40 60 80 100 120 Month Image data 2 1 Bayesian inference 2 Variational inference 3 Stochastic
More informationBayesian Inference and MCMC
Bayesian Inference and MCMC Aryan Arbabi Partly based on MCMC slides from CSC412 Fall 2018 1 / 18 Bayesian Inference - Motivation Consider we have a data set D = {x 1,..., x n }. E.g each x i can be the
More informationIntroduction to Machine Learning
Introduction to Machine Learning Brown University CSCI 1950-F, Spring 2012 Prof. Erik Sudderth Lecture 25: Markov Chain Monte Carlo (MCMC) Course Review and Advanced Topics Many figures courtesy Kevin
More informationParticle Filtering for Data-Driven Simulation and Optimization
Particle Filtering for Data-Driven Simulation and Optimization John R. Birge The University of Chicago Booth School of Business Includes joint work with Nicholas Polson. JRBirge INFORMS Phoenix, October
More informationLecture 8: Bayesian Estimation of Parameters in State Space Models
in State Space Models March 30, 2016 Contents 1 Bayesian estimation of parameters in state space models 2 Computational methods for parameter estimation 3 Practical parameter estimation in state space
More informationSequential Monte Carlo Methods for Bayesian Computation
Sequential Monte Carlo Methods for Bayesian Computation A. Doucet Kyoto Sept. 2012 A. Doucet (MLSS Sept. 2012) Sept. 2012 1 / 136 Motivating Example 1: Generic Bayesian Model Let X be a vector parameter
More informationLecture 13 : Variational Inference: Mean Field Approximation
10-708: Probabilistic Graphical Models 10-708, Spring 2017 Lecture 13 : Variational Inference: Mean Field Approximation Lecturer: Willie Neiswanger Scribes: Xupeng Tong, Minxing Liu 1 Problem Setup 1.1
More informationLecture 2: From Linear Regression to Kalman Filter and Beyond
Lecture 2: From Linear Regression to Kalman Filter and Beyond Department of Biomedical Engineering and Computational Science Aalto University January 26, 2012 Contents 1 Batch and Recursive Estimation
More informationBayesian Inference for Dirichlet-Multinomials
Bayesian Inference for Dirichlet-Multinomials Mark Johnson Macquarie University Sydney, Australia MLSS Summer School 1 / 50 Random variables and distributed according to notation A probability distribution
More informationLecture 4: Probabilistic Learning. Estimation Theory. Classification with Probability Distributions
DD2431 Autumn, 2014 1 2 3 Classification with Probability Distributions Estimation Theory Classification in the last lecture we assumed we new: P(y) Prior P(x y) Lielihood x2 x features y {ω 1,..., ω K
More informationAn Brief Overview of Particle Filtering
1 An Brief Overview of Particle Filtering Adam M. Johansen a.m.johansen@warwick.ac.uk www2.warwick.ac.uk/fac/sci/statistics/staff/academic/johansen/talks/ May 11th, 2010 Warwick University Centre for Systems
More informationCS281A/Stat241A Lecture 22
CS281A/Stat241A Lecture 22 p. 1/4 CS281A/Stat241A Lecture 22 Monte Carlo Methods Peter Bartlett CS281A/Stat241A Lecture 22 p. 2/4 Key ideas of this lecture Sampling in Bayesian methods: Predictive distribution
More informationRobotics. Mobile Robotics. Marc Toussaint U Stuttgart
Robotics Mobile Robotics State estimation, Bayes filter, odometry, particle filter, Kalman filter, SLAM, joint Bayes filter, EKF SLAM, particle SLAM, graph-based SLAM Marc Toussaint U Stuttgart DARPA Grand
More informationPATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 13: SEQUENTIAL DATA
PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 13: SEQUENTIAL DATA Contents in latter part Linear Dynamical Systems What is different from HMM? Kalman filter Its strength and limitation Particle Filter
More informationAn introduction to Sequential Monte Carlo
An introduction to Sequential Monte Carlo Thang Bui Jes Frellsen Department of Engineering University of Cambridge Research and Communication Club 6 February 2014 1 Sequential Monte Carlo (SMC) methods
More information27 : Distributed Monte Carlo Markov Chain. 1 Recap of MCMC and Naive Parallel Gibbs Sampling
10-708: Probabilistic Graphical Models 10-708, Spring 2014 27 : Distributed Monte Carlo Markov Chain Lecturer: Eric P. Xing Scribes: Pengtao Xie, Khoa Luu In this scribe, we are going to review the Parallel
More informationLecture 4: Probabilistic Learning
DD2431 Autumn, 2015 1 Maximum Likelihood Methods Maximum A Posteriori Methods Bayesian methods 2 Classification vs Clustering Heuristic Example: K-means Expectation Maximization 3 Maximum Likelihood Methods
More informationLinear Dynamical Systems
Linear Dynamical Systems Sargur N. srihari@cedar.buffalo.edu Machine Learning Course: http://www.cedar.buffalo.edu/~srihari/cse574/index.html Two Models Described by Same Graph Latent variables Observations
More informationMarkov Chain Monte Carlo Methods for Stochastic
Markov Chain Monte Carlo Methods for Stochastic Optimization i John R. Birge The University of Chicago Booth School of Business Joint work with Nicholas Polson, Chicago Booth. JRBirge U Florida, Nov 2013
More informationMachine Learning Summer School
Machine Learning Summer School Lecture 3: Learning parameters and structure Zoubin Ghahramani zoubin@eng.cam.ac.uk http://learning.eng.cam.ac.uk/zoubin/ Department of Engineering University of Cambridge,
More informationWeakness of Beta priors (or conjugate priors in general) They can only represent a limited range of prior beliefs. For example... There are no bimodal beta distributions (except when the modes are at 0
More informationState-Space Methods for Inferring Spike Trains from Calcium Imaging
State-Space Methods for Inferring Spike Trains from Calcium Imaging Joshua Vogelstein Johns Hopkins April 23, 2009 Joshua Vogelstein (Johns Hopkins) State-Space Calcium Imaging April 23, 2009 1 / 78 Outline
More informationBayesian Learning. HT2015: SC4 Statistical Data Mining and Machine Learning. Maximum Likelihood Principle. The Bayesian Learning Framework
HT5: SC4 Statistical Data Mining and Machine Learning Dino Sejdinovic Department of Statistics Oxford http://www.stats.ox.ac.uk/~sejdinov/sdmml.html Maximum Likelihood Principle A generative model for
More informationSTA 4273H: Statistical Machine Learning
STA 4273H: Statistical Machine Learning Russ Salakhutdinov Department of Statistics! rsalakhu@utstat.toronto.edu! http://www.utstat.utoronto.ca/~rsalakhu/ Sidney Smith Hall, Room 6002 Lecture 11 Project
More informationThe Particle Filter. PD Dr. Rudolph Triebel Computer Vision Group. Machine Learning for Computer Vision
The Particle Filter Non-parametric implementation of Bayes filter Represents the belief (posterior) random state samples. by a set of This representation is approximate. Can represent distributions that
More informationTSRT14: Sensor Fusion Lecture 8
TSRT14: Sensor Fusion Lecture 8 Particle filter theory Marginalized particle filter Gustaf Hendeby gustaf.hendeby@liu.se TSRT14 Lecture 8 Gustaf Hendeby Spring 2018 1 / 25 Le 8: particle filter theory,
More informationAuxiliary Particle Methods
Auxiliary Particle Methods Perspectives & Applications Adam M. Johansen 1 adam.johansen@bristol.ac.uk Oxford University Man Institute 29th May 2008 1 Collaborators include: Arnaud Doucet, Nick Whiteley
More informationLecture 7: Optimal Smoothing
Department of Biomedical Engineering and Computational Science Aalto University March 17, 2011 Contents 1 What is Optimal Smoothing? 2 Bayesian Optimal Smoothing Equations 3 Rauch-Tung-Striebel Smoother
More informationMarkov Chain Monte Carlo Methods for Stochastic Optimization
Markov Chain Monte Carlo Methods for Stochastic Optimization John R. Birge The University of Chicago Booth School of Business Joint work with Nicholas Polson, Chicago Booth. JRBirge U of Toronto, MIE,
More informationLecture : Probabilistic Machine Learning
Lecture : Probabilistic Machine Learning Riashat Islam Reasoning and Learning Lab McGill University September 11, 2018 ML : Many Methods with Many Links Modelling Views of Machine Learning Machine Learning
More informationLecture 6: Graphical Models: Learning
Lecture 6: Graphical Models: Learning 4F13: Machine Learning Zoubin Ghahramani and Carl Edward Rasmussen Department of Engineering, University of Cambridge February 3rd, 2010 Ghahramani & Rasmussen (CUED)
More informationMonte Carlo Methods. Geoff Gordon February 9, 2006
Monte Carlo Methods Geoff Gordon ggordon@cs.cmu.edu February 9, 2006 Numerical integration problem 5 4 3 f(x,y) 2 1 1 0 0.5 0 X 0.5 1 1 0.8 0.6 0.4 Y 0.2 0 0.2 0.4 0.6 0.8 1 x X f(x)dx Used for: function
More informationParticle Filtering Approaches for Dynamic Stochastic Optimization
Particle Filtering Approaches for Dynamic Stochastic Optimization John R. Birge The University of Chicago Booth School of Business Joint work with Nicholas Polson, Chicago Booth. JRBirge I-Sim Workshop,
More informationComputer Intensive Methods in Mathematical Statistics
Computer Intensive Methods in Mathematical Statistics Department of mathematics johawes@kth.se Lecture 7 Sequential Monte Carlo methods III 7 April 2017 Computer Intensive Methods (1) Plan of today s lecture
More informationIntegrated Non-Factorized Variational Inference
Integrated Non-Factorized Variational Inference Shaobo Han, Xuejun Liao and Lawrence Carin Duke University February 27, 2014 S. Han et al. Integrated Non-Factorized Variational Inference February 27, 2014
More informationECE276A: Sensing & Estimation in Robotics Lecture 10: Gaussian Mixture and Particle Filtering
ECE276A: Sensing & Estimation in Robotics Lecture 10: Gaussian Mixture and Particle Filtering Lecturer: Nikolay Atanasov: natanasov@ucsd.edu Teaching Assistants: Siwei Guo: s9guo@eng.ucsd.edu Anwesan Pal:
More informationBayesian Machine Learning
Bayesian Machine Learning Andrew Gordon Wilson ORIE 6741 Lecture 4 Occam s Razor, Model Construction, and Directed Graphical Models https://people.orie.cornell.edu/andrew/orie6741 Cornell University September
More informationIntroduction to Particle Filters for Data Assimilation
Introduction to Particle Filters for Data Assimilation Mike Dowd Dept of Mathematics & Statistics (and Dept of Oceanography Dalhousie University, Halifax, Canada STATMOS Summer School in Data Assimila5on,
More informationLecture 6: Bayesian Inference in SDE Models
Lecture 6: Bayesian Inference in SDE Models Bayesian Filtering and Smoothing Point of View Simo Särkkä Aalto University Simo Särkkä (Aalto) Lecture 6: Bayesian Inference in SDEs 1 / 45 Contents 1 SDEs
More informationApproximate Bayesian Computation and Particle Filters
Approximate Bayesian Computation and Particle Filters Dennis Prangle Reading University 5th February 2014 Introduction Talk is mostly a literature review A few comments on my own ongoing research See Jasra
More informationLecture 2: Conjugate priors
(Spring ʼ) Lecture : Conjugate priors Julia Hockenmaier juliahmr@illinois.edu Siebel Center http://www.cs.uiuc.edu/class/sp/cs98jhm The binomial distribution If p is the probability of heads, the probability
More informationSTA 4273H: Sta-s-cal Machine Learning
STA 4273H: Sta-s-cal Machine Learning Russ Salakhutdinov Department of Computer Science! Department of Statistical Sciences! rsalakhu@cs.toronto.edu! h0p://www.cs.utoronto.ca/~rsalakhu/ Lecture 2 In our
More informationComputer Intensive Methods in Mathematical Statistics
Computer Intensive Methods in Mathematical Statistics Department of mathematics johawes@kth.se Lecture 16 Advanced topics in computational statistics 18 May 2017 Computer Intensive Methods (1) Plan of
More informationIntroduction to Bayesian Methods
Introduction to Bayesian Methods Jessi Cisewski Department of Statistics Yale University Sagan Summer Workshop 2016 Our goal: introduction to Bayesian methods Likelihoods Priors: conjugate priors, non-informative
More informationGaussian processes. Chuong B. Do (updated by Honglak Lee) November 22, 2008
Gaussian processes Chuong B Do (updated by Honglak Lee) November 22, 2008 Many of the classical machine learning algorithms that we talked about during the first half of this course fit the following pattern:
More informationGraphical Models and Kernel Methods
Graphical Models and Kernel Methods Jerry Zhu Department of Computer Sciences University of Wisconsin Madison, USA MLSS June 17, 2014 1 / 123 Outline Graphical Models Probabilistic Inference Directed vs.
More informationCSC2535: Computation in Neural Networks Lecture 7: Variational Bayesian Learning & Model Selection
CSC2535: Computation in Neural Networks Lecture 7: Variational Bayesian Learning & Model Selection (non-examinable material) Matthew J. Beal February 27, 2004 www.variational-bayes.org Bayesian Model Selection
More informationIntroduction to Gaussian Processes
Introduction to Gaussian Processes Iain Murray murray@cs.toronto.edu CSC255, Introduction to Machine Learning, Fall 28 Dept. Computer Science, University of Toronto The problem Learn scalar function of
More informationSequential Monte Carlo Methods (for DSGE Models)
Sequential Monte Carlo Methods (for DSGE Models) Frank Schorfheide University of Pennsylvania, PIER, CEPR, and NBER October 23, 2017 Some References These lectures use material from our joint work: Tempered
More informationBayesian Estimation with Sparse Grids
Bayesian Estimation with Sparse Grids Kenneth L. Judd and Thomas M. Mertens Institute on Computational Economics August 7, 27 / 48 Outline Introduction 2 Sparse grids Construction Integration with sparse
More informationExercises Tutorial at ICASSP 2016 Learning Nonlinear Dynamical Models Using Particle Filters
Exercises Tutorial at ICASSP 216 Learning Nonlinear Dynamical Models Using Particle Filters Andreas Svensson, Johan Dahlin and Thomas B. Schön March 18, 216 Good luck! 1 [Bootstrap particle filter for
More informationEfficient Variational Inference in Large-Scale Bayesian Compressed Sensing
Efficient Variational Inference in Large-Scale Bayesian Compressed Sensing George Papandreou and Alan Yuille Department of Statistics University of California, Los Angeles ICCV Workshop on Information
More informationPMR Learning as Inference
Outline PMR Learning as Inference Probabilistic Modelling and Reasoning Amos Storkey Modelling 2 The Exponential Family 3 Bayesian Sets School of Informatics, University of Edinburgh Amos Storkey PMR Learning
More informationDoing Bayesian Integrals
ASTR509-13 Doing Bayesian Integrals The Reverend Thomas Bayes (c.1702 1761) Philosopher, theologian, mathematician Presbyterian (non-conformist) minister Tunbridge Wells, UK Elected FRS, perhaps due to
More informationBayesian Regression Linear and Logistic Regression
When we want more than point estimates Bayesian Regression Linear and Logistic Regression Nicole Beckage Ordinary Least Squares Regression and Lasso Regression return only point estimates But what if we
More informationSTA 4273H: Statistical Machine Learning
STA 4273H: Statistical Machine Learning Russ Salakhutdinov Department of Statistics! rsalakhu@utstat.toronto.edu! http://www.utstat.utoronto.ca/~rsalakhu/ Sidney Smith Hall, Room 6002 Lecture 7 Approximate
More informationComputational Cognitive Science
Computational Cognitive Science Lecture 8: Frank Keller School of Informatics University of Edinburgh keller@inf.ed.ac.uk Based on slides by Sharon Goldwater October 14, 2016 Frank Keller Computational
More informationLECTURE 15 Markov chain Monte Carlo
LECTURE 15 Markov chain Monte Carlo There are many settings when posterior computation is a challenge in that one does not have a closed form expression for the posterior distribution. Markov chain Monte
More informationLinear Models A linear model is defined by the expression
Linear Models A linear model is defined by the expression x = F β + ɛ. where x = (x 1, x 2,..., x n ) is vector of size n usually known as the response vector. β = (β 1, β 2,..., β p ) is the transpose
More informationExpectation Propagation Algorithm
Expectation Propagation Algorithm 1 Shuang Wang School of Electrical and Computer Engineering University of Oklahoma, Tulsa, OK, 74135 Email: {shuangwang}@ou.edu This note contains three parts. First,
More informationLecture 13 Fundamentals of Bayesian Inference
Lecture 13 Fundamentals of Bayesian Inference Dennis Sun Stats 253 August 11, 2014 Outline of Lecture 1 Bayesian Models 2 Modeling Correlations Using Bayes 3 The Universal Algorithm 4 BUGS 5 Wrapping Up
More informationThe Kalman Filter ImPr Talk
The Kalman Filter ImPr Talk Ged Ridgway Centre for Medical Image Computing November, 2006 Outline What is the Kalman Filter? State Space Models Kalman Filter Overview Bayesian Updating of Estimates Kalman
More informationCalibration of Stochastic Volatility Models using Particle Markov Chain Monte Carlo Methods
Calibration of Stochastic Volatility Models using Particle Markov Chain Monte Carlo Methods Jonas Hallgren 1 1 Department of Mathematics KTH Royal Institute of Technology Stockholm, Sweden BFS 2012 June
More informationParameter estimation and forecasting. Cristiano Porciani AIfA, Uni-Bonn
Parameter estimation and forecasting Cristiano Porciani AIfA, Uni-Bonn Questions? C. Porciani Estimation & forecasting 2 Temperature fluctuations Variance at multipole l (angle ~180o/l) C. Porciani Estimation
More informationProbabilistic Graphical Models
Probabilistic Graphical Models Lecture 11 CRFs, Exponential Family CS/CNS/EE 155 Andreas Krause Announcements Homework 2 due today Project milestones due next Monday (Nov 9) About half the work should
More information13: Variational inference II
10-708: Probabilistic Graphical Models, Spring 2015 13: Variational inference II Lecturer: Eric P. Xing Scribes: Ronghuo Zheng, Zhiting Hu, Yuntian Deng 1 Introduction We started to talk about variational
More informationData Analysis and Uncertainty Part 2: Estimation
Data Analysis and Uncertainty Part 2: Estimation Instructor: Sargur N. University at Buffalo The State University of New York srihari@cedar.buffalo.edu 1 Topics in Estimation 1. Estimation 2. Desirable
More informationDAG models and Markov Chain Monte Carlo methods a short overview
DAG models and Markov Chain Monte Carlo methods a short overview Søren Højsgaard Institute of Genetics and Biotechnology University of Aarhus August 18, 2008 Printed: August 18, 2008 File: DAGMC-Lecture.tex
More informationBayesian Methods in Positioning Applications
Bayesian Methods in Positioning Applications Vedran Dizdarević v.dizdarevic@tugraz.at Graz University of Technology, Austria 24. May 2006 Bayesian Methods in Positioning Applications p.1/21 Outline Problem
More information9 Forward-backward algorithm, sum-product on factor graphs
Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science 6.438 Algorithms For Inference Fall 2014 9 Forward-backward algorithm, sum-product on factor graphs The previous
More informationMarkov Chains and Hidden Markov Models
Chapter 1 Markov Chains and Hidden Markov Models In this chapter, we will introduce the concept of Markov chains, and show how Markov chains can be used to model signals using structures such as hidden
More informationTopic Models. Charles Elkan November 20, 2008
Topic Models Charles Elan elan@cs.ucsd.edu November 20, 2008 Suppose that we have a collection of documents, and we want to find an organization for these, i.e. we want to do unsupervised learning. One
More informationIntroduction into Bayesian statistics
Introduction into Bayesian statistics Maxim Kochurov EF MSU November 15, 2016 Maxim Kochurov Introduction into Bayesian statistics EF MSU 1 / 7 Content 1 Framework Notations 2 Difference Bayesians vs Frequentists
More informationLearning of dynamical systems
Learning of dynamical systems Particle filters and Markov chain methods Thomas B. Schön and Fredrik Lindsten c Draft date August 23, 2017 2 Contents 1 Introduction 3 1.1 A few words for readers of the
More informationNonparametric Bayesian Methods (Gaussian Processes)
[70240413 Statistical Machine Learning, Spring, 2015] Nonparametric Bayesian Methods (Gaussian Processes) Jun Zhu dcszj@mail.tsinghua.edu.cn http://bigml.cs.tsinghua.edu.cn/~jun State Key Lab of Intelligent
More informationMCMC 2: Lecture 3 SIR models - more topics. Phil O Neill Theo Kypraios School of Mathematical Sciences University of Nottingham
MCMC 2: Lecture 3 SIR models - more topics Phil O Neill Theo Kypraios School of Mathematical Sciences University of Nottingham Contents 1. What can be estimated? 2. Reparameterisation 3. Marginalisation
More information(5) Multi-parameter models - Gibbs sampling. ST440/540: Applied Bayesian Analysis
Summarizing a posterior Given the data and prior the posterior is determined Summarizing the posterior gives parameter estimates, intervals, and hypothesis tests Most of these computations are integrals
More informationBayesian Inference for DSGE Models. Lawrence J. Christiano
Bayesian Inference for DSGE Models Lawrence J. Christiano Outline State space-observer form. convenient for model estimation and many other things. Bayesian inference Bayes rule. Monte Carlo integation.
More informationPrinciples of Bayesian Inference
Principles of Bayesian Inference Sudipto Banerjee University of Minnesota July 20th, 2008 1 Bayesian Principles Classical statistics: model parameters are fixed and unknown. A Bayesian thinks of parameters
More informationBayesian Machine Learning
Bayesian Machine Learning Andrew Gordon Wilson ORIE 6741 Lecture 2: Bayesian Basics https://people.orie.cornell.edu/andrew/orie6741 Cornell University August 25, 2016 1 / 17 Canonical Machine Learning
More informationKalman filtering and friends: Inference in time series models. Herke van Hoof slides mostly by Michael Rubinstein
Kalman filtering and friends: Inference in time series models Herke van Hoof slides mostly by Michael Rubinstein Problem overview Goal Estimate most probable state at time k using measurement up to time
More informationLecture 7 and 8: Markov Chain Monte Carlo
Lecture 7 and 8: Markov Chain Monte Carlo 4F13: Machine Learning Zoubin Ghahramani and Carl Edward Rasmussen Department of Engineering University of Cambridge http://mlg.eng.cam.ac.uk/teaching/4f13/ Ghahramani
More informationWeakness of Beta priors (or conjugate priors in general) They can only represent a limited range of prior beliefs. For example... There are no bimodal beta distributions (except when the modes are at 0
More informationMarkov chain Monte Carlo Lecture 9
Markov chain Monte Carlo Lecture 9 David Sontag New York University Slides adapted from Eric Xing and Qirong Ho (CMU) Limitations of Monte Carlo Direct (unconditional) sampling Hard to get rare events
More information