Hidden Markov Model

Reprinted from Zhihu

Author: Fatty Wang
Link: https://www.zhihu.com/question/20962240/answer/33614574

 

 

 

Simple explanation:

I am a fighter with three combat forms, namely, rage state , normal state and defensive state . At the same time, I also have three passive skills, which are normal flat A, critical strike (double the attack damage), and blood drain (life drain).
In the rage state, the probability of making a critical strike is 80%, and the probability of making a lifesteal is 5%;
in the balanced state, the probability of making a critical strike is 30%, and the probability of making a lifesteal is 20%;
in the defensive form, the critical strike The success probability is 5%, and the lifesteal probability is 60%.

A Hidden Markov (HMM) model is such a system - it has a hidden state that changes over time, continuously affecting its appearance
Markov chain: The latter state is only related to the previous state, and has nothing to do with the previous one.
              For example: Tomorrow's result is only related to today's, and has nothing to do with yesterday's

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=326263975&siteId=291194637