Efficient EM Learning with Tabulation for Parameterized Logic Programs
Efficient EM Learning with Tabulation for Parameterized Logic Programs
复制标题
通过参数化逻辑程序的表格进行高效 EM 学习
DOI：
10.1007/3-540-44957-4_18
复制
发表时间：
2000
期刊：
影响因子：
--
通讯作者：
Taisuke Sato
中科院分区：
文献类型：
--
作者：
Yoshitaka Kameya;Taisuke Sato
我们一直在开发一种基于逻辑编程框架的通用符号统计建模语言[6,19,20]，该框架在语义上统一（并扩展）主要符号统计框架，例如隐马尔可夫模型（HMM）[18]、概率上下文无关语法（PCFG）[23]和贝叶斯网络[16]。 PRISM 语言旨在对受基于分布语义的规则和概率控制的复杂符号现象进行建模[19]。程序包含统计参数，它们是通过专门派生的 EM 算法（图形 EM 算法）从随机采样的数据中自动学习的。它适用于表示观察目标解释的共享结构的支持图。在本文中，我们提出使用制表技术来构建支持图，并表明，图形 EM 算法的时间复杂度与 HMM（Baum-Welch 算法 [18]）和 PCFG（Inside-Outside 算法 [1]）的专用 EM 算法相同。
We have been developing a general symbolic-statistical modeling language [6,19,20] based on the logic programming framework that semantically unifies (and extends) major symbolic-statistical frameworks such as hidden Markov models (HMMs) [18], probabilistic context-free grammars (PCFGs) [23] and Bayesian networks [16]. The language, PRISM, is intended to model complex symbolic phenomena governed by rules and probabilities based on the distributional semantics[19]. Programs contain statistical parameters and they are automatically learned from randomly sampled data by a specially derived EM algorithm, the graphical EM algorithm. It works on support graphs representing the shared structure of explanations for an observed goal. In this paper, we propose the use of tabulation technique to build support graphs, and show that as a result, the graphical EM algorithm attains the same time complexity as specilized EM algorithms for HMMs (the Baum-Welch algorithm [18]) and PCFGs (the Inside-Outside algorithm [1]).