machine-learning - 向前向后隐藏马尔可夫模型与左右向隐藏马尔可夫模型有什么区别吗？(Is there any Difference between forward backward hidden markov model and left-right hidden markov model?)

Question

向前向后隐藏马尔可夫模型与左右向隐藏马尔可夫模型之间有什么区别吗？

  ask by Riaz Ali Hunzai translate from so

Answer 1

forward and backward is how you calculate some probabilities of interest using ready to use hmm models and definitions.

(前进和后退是如何使用随时可用的hmm模型和定义来计算某些感兴趣的概率。)

left to right is correspoding to structure of underlying hidden state of Markov Chain for an HMM.

(从左到右对应于HMM的马尔可夫链的基础隐藏状态的结构。)

This is because HMM can be viewed as a Markov Chain that also generate observations.

(这是因为HMM可被视为也可产生观测值的马尔可夫链。)

so left to right is nothing more than left to right Markov Chain, a MC that only can go to next other state and never come back to previous states again as time propagates.

(因此从左到右无非是从左到右的马尔可夫链，这是一个MC，它只能进入下一个其他状态，而随着时间的流逝再也不会回到先前的状态。)