N ov 2 02 0 DYNAMIC INFERENCE IN PROBABILISTIC GRAPHICAL MODELS
Weiming Feng,Xiaoming Sun,Yitong Yin
2021-01-01
Abstract:Probabilistic graphical models, such as Markov random fields (MRFs), are useful for describing high-dimensional distributions in terms of local dependence structures. e probabilistic inference is a fundamental problem related to graphical models, and sampling is a main approach for the problem. In this paper, we study probabilistic inference problems when the graphical model itself is changing dynamically with time. Such dynamic inference problems arise naturally in today’s application, e.g. multivariate time-series data analysis and practical learning procedures. We give a dynamic algorithm for sampling-based probabilistic inferences in MRFs, where each dynamic update can change the underlying graph and all parameters of the MRF simultaneously, as long as the total amount of changes is bounded. More precisely, suppose that the MRF has = variables and polylogarithmic-bounded maximum degree, and # (=) independent samples are sufficient for the inference for a polynomial function # (·). Our algorithm dynamically maintains an answer to the inference problem using $̃ (=# (=)) space cost, and $̃ (# (=)+=) incremental time cost upon each update to theMRF, as long as the Dobrushin-Shlosman condition is satisfied by the MRFs. is well-known condition has long been used for guaranteeing the efficiency of Markov chain Monte Carlo (MCMC) sampling in the traditional static seing. Compared to the static case, which requires Ω(=# (=)) time cost for redrawing all # (=) samples whenever the MRF changes, our dynamic algorithm gives a Ω̃(min{=,# (=)})-factor speedup. Our approach relies on a novel dynamic sampling technique, which transforms local Markov chains (a.k.a. single-site dynamics) to dynamic sampling algorithms, and an “algorithmic Lipschitz” condition that we establish for sampling from graphical models, namely, when the MRF changes by a small difference, samples can bemodified to reflect the new distribution, with cost proportional to the difference on MRF. (Weiming Feng, Yitong Yin) State Key Laboratory for Novel Software Technology, Nanjing University. E-mail: fengwm@smail.nju.edu.cn,yinyt@nju.edu.cn (Kun He) Shenzhen University; Shenzhen Institute of Computing Sciences. E-mail: hekun.threebody@foxmail.com (Xiaoming Sun) CAS Key Lab of Network Data Science and Technology, Institute of Computing Technology, Chinese Academy of Sciences. E-mail: sunxiaoming@ict.ac.cn Weiming Feng and Yitong Yin are supported by theNational Key R&DProgram of China 2018YFB1003202and theNational Natural Science Foundation of China under Grant Nos. 61722207 and 61672275. Kun He and Xiaoming Sun are supported by the National Natural Science Foundation of China Grants No. 61832003, 61433014 and K.C.Wong Education Foundation.