翻訳と辞書 Words near each other ・ Variegated lizardfish ・ Variegated meadowhawk ・ Variegated mogurnda ・ Variegated mountain lizard ・ Variegated pink lemon ・ Variegated squirrel ・ Variegated tinamou ・ Variegated tree frog ・ Variegated wolf ・ Variegated yarn ・ Variegatic acid ・ Variegation ・ Variational bicomplex ・ Variational inequality ・ Variational integrator ・ Variational message passing ・ Variational method (quantum mechanics) ・ Variational methods in general relativity ・ Variational Monte Carlo ・ Variational perturbation theory ・ Variational principle ・ Variational properties ・ Variational transition-state theory ・ Variational vector field ・ Variations (Andrew Lloyd Webber album) ・ Variations (ballet) ・ Variations (Cage) ・ Variations (Eddie Rabbitt album) ・ Variations (Stravinsky) ・ Variations and drugs database Dictionary Lists mini英和辞書 mini和英辞書 Webster 1913 Latin-English FOLDOC Wikipedia English ウィキペディア

翻訳と辞書　辞書検索 [ 開発暫定版 ] スポンサード リンク Variational message passing ： ウィキペディア英語版 Variational message passing Variational message passing (VMP) is an approximate inference technique for continuous- or discrete-valued Bayesian networks, with conjugate-exponential parents, developed by John Winn. VMP was developed as a means of generalizing the approximate variational methods used by such techniques as Latent Dirichlet allocation and works by updating an approximate distribution at each node through messages in the node's Markov blanket. ==Likelihood Lower Bound== Given some set of hidden variables $H$ and observed variables $V$, the goal of approximate inference is to lower-bound the probability that a graphical model is in the configuration $V$. Over some probability distribution $Q$ (to be defined later), :$\backslash ln\; P(V)\; =\; \backslash sum\_H\; Q(H)\; \backslash ln\; \backslash frac\; =\; \backslash sum\_\; Q(H)\; \backslash Bigg(\backslash ln\; \backslash frac\; -\; \backslash ln\; \backslash frac\; \backslash Bigg)$. So, if we define our lower bound to be :$L(Q)\; =\; \backslash sum\_\; Q(H)\; \backslash ln\; \backslash frac$, then the likelihood is simply this bound plus the relative entropy between $P$ and $Q$. Because the relative entropy is non-negative, the function $L$ defined above is indeed a lower bound of the log likelihood of our observation $V$. The distribution $Q$ will have a simpler character than that of $P$ because marginalizing over $P$ is intractable for all but the simplest of graphical models. In particular, VMP uses a factorized distribution $Q$: :$Q(H)\; =\; \backslash prod\_i\; Q\_i(H\_i),$ where $H\_i$ is a disjoint part of the graphical model. 抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』 ■ウィキペディアで「Variational message passing」の詳細全文を読む スポンサード リンク 翻訳と辞書 : 翻訳のためのインターネットリソース Copyright(C) kotoba.ne.jp 1997-2016. All Rights Reserved.