- -
UPV
 

Forward and Backward algorithms

Juan Císcar, Alfonso; Sanchis Navarro, José Alberto; Civera Saiz, Jorge; The training objectives of the learning object are: 1) To explain the difficulty of computing the probability of a string with a Hidden Markov Model (HMM); 2) To compute the prob. of a string with the Forward algorithm; and 3) To compute the prob. of a string with the Backward algorithm. In this regard, it is worth noting that, given an HMM, both the Forward and Backward algorithms are commonly used for efficient computation of the exact probability of a string. In this learning object, these algorithms are described at a basic level with the help of simple examples. http://hdl.handle.net/10251/122821


EMAS upv