GKI Inference in Bayesian Networks 作业笔记

这篇博客探讨了马尔科夫链在天气预报中的应用,以及隐藏马尔科夫模型在安全监控场景中的概率计算。通过一系列问题,解释了如何建立贝叶斯网络、转移图,并使用过滤和维特比算法来推断最可能的状态序列。
摘要由CSDN通过智能技术生成
  • Markov Chains Markov-Ketten
    The weather of a day depends only on the previous day.
    question 1: probability, that it rains in 2 days(两天以后),if it has rained today?
    question 2: probability, P ( R 1 = f ∣ R 2 = t , R 0 = t ) P(R_1=f|R_2=t, R_0=t) P(R1=fR2=t,R0=t)

  • Hidden Markov Models
    you work as a security guard in a secret underground facility. You judge whether it is a rainy day by whether your director carries an umbrella. P ( R 0 = t ) = 0.5 P(R_0=t)=0.5 P(R0=t)=0.5 ist given in advance.
    question 1: draw a bayesian network with a starting state and n mor time states. Also draw the transitin diagram.
    question 2: P ( R 1 = t ∣ U 1 = f ) P(R_1 = t| U_1 = f) P(R1=tU1=f)
    use filtering
    question 3: P ( R 1 = t ∣ U 1 = f , U 2 = t ) P(R_1=t|U_1=f, U_2=t) P(R1=tU1=f,U2=t)
    a future observation is included in the estimation of R 1 R_1 R1 => smoothing
    Filter distributions P ( R 1 ∣ U 1 ) P(R_1 | U_1) P(R1U1) have already been calculated in question 1
    backward calculation, P ( U 2 ∣ R 1 ) P(U_2|R_1) P(U2R1)
    P ( X t ∣ Y 1 , . . . , Y T ) = α P ( Y t + 1 , . . . , Y T ∣ X t ) P ( X t ∣ Y 1

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值