Right now you can run this to try multiple options for predicting possibilities of seeing an observation when you know the states that the agent goes through in the context of spelling. You can also set the parameters of hitting the correct letter you intend to type, and how far you deviate if you are to miss.
royalsix / spellingrecognitionhmm Goto Github PK
View Code? Open in Web Editor NEWSimple Hidden Markov Model For Spelling Recognition