Bayesian Network Exercise 01

April 6, 2012, Tags: AI

Introduction

AI Course Bayesian network excerice 01 resolution using enumeration method.

  • C 国家政策
  • U 学校政策
  • W 工作压力
  • B 身体状况
  • D 过劳死

Scenarios

  • 已知 : Exercise01
  • 求 : 当 学校政策==True 而且 工作压力 == True 时, 过劳死的概率.

Resolution

  • Value T:0.95 means True case with probability 0.95
  • Probability of C is omitted as it is not related. Even include C, the result should be same.
  • Enumeration list

    INDEX U W B D
    1 T:0.96 T:0.90 T:0.3 T:0.335
    2 T:0.96 T:0.90 F:0.7 T:0.3
    3 T:0.96 T:0.90 T:0.3 F:0.665
    4 T:0.06 T:0.90 F:0.7 F:0.7
  • Probability calculation


P(D=T|U=T,W=T)
  = ( P(1) + P(2) ) / ( P(1) + P(2 )+ ... + P(4) )
  = (0.96*0.90*0.3*0.335 + 0.96*0.90*0.7*0.3 ) / (...)
  = 0.3 * 1.035 / (0.3 * 1.035 + 0.3 * 0.665 + 0.7 * 0.7)
  = 31.05%