摘要: $P((y \Rightarrow x) | x, y) = \prod\limits_k(1-(1-y_k)x_k)$ 阅读全文
posted @ 2016-08-05 15:10 林大勇 阅读(211) 评论(0) 推荐(0) 编辑
摘要: $q$(S $\rightarrow$ NP VP) * $q$(NP $\rightarrow$ NNP) * $q$(VP $\rightarrow$ VB NP) * $q$(NP $\rightarrow$ NNP)* $q$(NNP $\rightarrow$ $IBM$) * $q$(V... 阅读全文
posted @ 2016-01-13 12:40 林大勇 阅读(175) 评论(0) 推荐(0) 编辑
摘要: 1. 前言:本文主要参考Alchemy Tutorial, washington主页上挂出的所有Alchemy项目(包括Alchemy1.0, Alchemy2.0, AlchemyLite)都无法在Ubuntu15.04中安装使用;这篇博文使用的Alchemy是PhDP大神在Github上的Alc... 阅读全文
posted @ 2015-12-23 10:49 林大勇 阅读(347) 评论(0) 推荐(0) 编辑
摘要: Dear Professor: I am a university student in China and study on MLN recently; I have a question as follow: The joint distribution repre... 阅读全文
posted @ 2015-12-11 17:57 林大勇 阅读(265) 评论(0) 推荐(0) 编辑
摘要: MLE & MAP : data / model parameterMLE: (1) keep the data fixed(i.e., it has been observed) and allow the parameters to vary (2) the likelihood func... 阅读全文
posted @ 2015-12-02 16:05 林大勇 阅读(277) 评论(0) 推荐(0) 编辑
摘要: If $x$ is a possible world (relational database) and $x_l$ is the $l$th ground atom's truth value,the pseudo-log-likelihood of $x$ given weights $w$ i... 阅读全文
posted @ 2015-11-24 12:11 林大勇 阅读(177) 评论(0) 推荐(0) 编辑
摘要: which mlnlearn:/home/panjinbo/Desktop/MLN/probcog/Toolbox/ProbCog-1.1/apps/mlnlearn==>/home/panjinbo/Desktop/MLN/probcog/Toolbox/ProbCog-1.1/python/ml... 阅读全文
posted @ 2015-11-24 11:35 林大勇 阅读(288) 评论(0) 推荐(0) 编辑
摘要: We consider two types of inference:finding the most likely state of the world consistent with some evidencecomputing arbitrary conditional probabiliti... 阅读全文
posted @ 2015-11-12 09:47 林大勇 阅读(599) 评论(0) 推荐(0) 编辑
摘要: MathJax与LaTex公式简介 (转载) PS: 原文链接写的非常好!!! 博主写这篇文章,一是为了防止原链接失效,二是在cnblogs上测试MathJax; 本文从math.stackexchange.com上名为MathJax basic tutorial and quick referen 阅读全文
posted @ 2015-11-11 11:15 林大勇 阅读(32978) 评论(2) 推荐(16) 编辑
摘要: 一. MLN相关知识的介绍1. First-order logicA first-order logic knowledge base (KB) is a set of formulas in first-order logic;Predicate symbols represent relatio... 阅读全文
posted @ 2015-10-30 11:56 林大勇 阅读(720) 评论(0) 推荐(0) 编辑