zoukankan      html  css  js  c++  java
  • Pedro domingos

    Dear Professor:

              I am a university student in China and study on MLN recently;
             
              I have a question as follow:
              The joint distribution represented by a Markov network is given by $$P(X=x) = frac{1}{Z} prod_k phi_{k}(x_{{k}}) ag{2.1}$$ $$P(X=x) = frac{1}{Z} expleft(sum_j w_jf_j(x) ight) ag{2.2}$$
              The probability distribution over possible worlds x specified by the ground Markov network $M_{L, C}$ is given by $$P(X=x) = frac{1}{Z} expleft(sum_i w_in_i(x) ight) = frac{1}{Z}prod_i phi_{i}(x_{{i}})^{n_i(x)} ag{2.3}$$
              The probability of a ground predicate $X_l$ when its Markov blanket $B_l$ is in state $b_l$ is $$P(X_l=x_l|B_l=b_l) = frac{exp(sum_{f_i in F_l} w_if_i(X_l=x_l, B_l=b_l)))}{exp(sum_{f_i in F_l} w_if_i(X_l=0, B_l=b_l))+exp(sum_{f_i in F_l} w_if_i(X_l=1, B_l=b_l))} ag{3.3}$$
         
              I can't figure out how to derive equation(3.3) from equation(2.1) or equation(2.3), maybe equation(3.3) is a definition, not a derivation ??
              
  • 相关阅读:
    sessionStorage用于分页,瀑布流和存储用户数据等
    js瀑布流
    sql 日结
    js 去除html标签
    c# 去除文本的html标签
    jQuery 数据滚动(上下)
    jQuery 图片随滚动条滚动加载
    sql 指定范围 获取随机数
    js 时间格式化
    js自写字符串 append 方法
  • 原文地址:https://www.cnblogs.com/linxd/p/5039801.html
Copyright © 2011-2022 走看看