yxnchen / PFA

Implementations of various Poisson Factor Analysis (PFA) models

Poisson Factor Analysis

Implementations of various Poisson Factor Analysis (PFA) models.

File Description

Model

• BHPF.m : Implements Gibbs sampling for Bayesian Hierarchical Poisson Factorization (BHPF).
• BPTF.m : Implements Gibbs sampling for Bayesian Poisson Tensor Factorization (BPTF).
• BGGP.m :
• GaP_v1.m :
• GaP_v2.m :
• GaP_v3.m :
• HaLRTC.m : A baseline model using low rank tensor completion (HaLRTC).

Tool

• kr.m :
• mat2ten.m :
• ten2mat.m :
• cp_combination.m :
• missing_setting.m :

Running

• main_code.m : The main code file. Arguments low_rank and missing_rate are supposed to be modified.

Dataset

• tensor.mat : See Urban Traffic Speed Dataset of Guangzhou.