Active Learning with Bayesian Nonnegative MatrixFactorization for Recommender Systems

Contributors:

• Gönül Aycı
• Abdullatif Köksal
• Melih Mutlu
• Burak Suyunu
• Prof. Ali Taylan Cemgil

Dense Dataset

	KL Divergence	RMSE
Random
Random Popularity
Max Row-Column
Max Variance
Min Variance
Popularity Variance

Sparse Dataset

	KL Divergence	RMSE
Random
Random Popularity
Max Row-Column
Max Variance
Min Variance
Popularity Variance