人工智能的课件CH14_Bayesian network-part2.pptVIP

Likelihood weighting example Likelihood weighting analysis Hence likelihood weighting returns consistent estimates but performance still degrades with many evidence variables because a few samples have nearly all the total weight Approximate inference using MCMC “State” of network = current assignment to all variables. Generate next state by sampling one variable given Markov blanket Sample each variable in turn, keeping evidence fixed Can also choose a variable to sample at random each time The Markov chain Wander about for a while, average what you see MCMC example contd. Theorem: chain approaches stationary distribution: long-run fraction of time spent in each state is exactly proportional to its posterior probability Markov blanket sampling Summary Exact inference by variable elimination: polytime on polytrees, NP-hard on general graphs space = time, very sensitive to topology Approximate inference by LW, MCMC: LW does poorly when there is lots of (downstream) evidence LW, MCMC generally insensitive to topology Convergence can be very slow with probabilities close to 1 or 0 Can handle arbitrary combinations of discrete and continuous variables Assignments Ex 14.3 * CPT=conditional probability table. n个d种取值变量. * SAT:SAT的一个实例是k个布尔变量1 x ，…， k x 的m个布尔表达式1 A ，…， m A 。若存在 各布尔变量i x (1≤i≤k)的0，1 赋值，使每个布尔表达式i A (1≤i≤m)都取值1，则称布尔表 达式1 A 2 A … m A 是可满足的。 k-SAT:如果一个布尔合取范式的每个乘积项最多是k个因子的析取式，就称之为k元合取范式， 简记为k-CNF。一个k-SAT问题是判定一个k-CNF是否可满足。特别地，当k=3 时，3-SAT 问题在NP完全问题树中具有重要地位。 * 智能计算研究中心 XIV. Bayesian networks(section 4-5) Autumn 2012 Instructor: Wang Xiaolong Harbin Institute of Technology, Shenzhen Graduate School Intelligent Computation Research Center (HITSGS ICRC) Outlines Exact inference by enumeration Exact inference by variable elimination Approximate inference by stochastic simulation Approximate inference by Markov chain Monte Carlo Inference tasks Simple queries: compute posterior marginal Conjunctive queries: Optimal decisions: decision network