theano-BPR
theano-BPR copied to clipboard
Using theano to implement the Matrix Factorization with BPR ranking loss
theano-BPR
============ Using Theano to implement the matrix factorization with BPR ranking loss, as described in:
Steffen Rendle, et al. BPR: Bayesian personalized ranking from implicit feedback UAI'09
Optimizer: Batched SGD
Usage:
Required softwares: numpy, theano
$ python run_example.py
Evaluation:
TopK recommendation (hold-1-out evaluation), with measures Hit Ratio and NDCG. More details about the evaluation can be found in:
Xiangnan He, et al. Fast Matrix Factorization for Online Recommendation with Implicit Feedback SIGIR'16
Welcome any comments for improving the efficency!
Author:
Xiangnan He (http://www.comp.nus.edu.sg/~xiangnan/)
Contact: [email protected]