---

Name: A Hidden Markov Model Purpose: A class for the model, including parameters. Author: Haidong Zhang Created: 28/08/2014 E-mail: [email protected]

DataPreprocess.py (Firstly excute.)

Pre-process the data, generate two .txt file. 'artist.txt': artist_id: the artist id. num: the number of artists listened to.

'users_artists_timestamp.txt': user_id: the user id. artist_id: the artist id. timestamp: the time stamp.

---

HMMForCF A Hidden Markov Model for Collaborative Filtering

---

HMM.py

A class for modeling the hidden markov model. HiddenStatesNum : The number of hidden states. ObservationStatesNum: The number of observation states. InitProbs: The probability of initial vector. TransProbs: he probability of transition matrix. Theta: The probability of emission matrix. a, b: Represent the parameters of negative binomial distribution.

---

HMM_ForwardBackward1.py

Forward-backward algorithm is used to train the model.

---

CommonFunc.py

Some common used functions.

---

UserModel.py

Model the users.

---

.