원문정보
초록
영어
The matrix factorization algorithms such as the matrix factorization technique (MF), singular value decomposition (SVD) and the probability matrix factorization (PMF) and so on, are summarized and compared. Based on the above research work, a kind of improved probability matrix factorization algorithm called MPMF is proposed in this paper. MPMF determines the optimal value of dimension D of both the user feature vector and the item feature vector through experiments. The complexity of the algorithm scales linearly with the number of observations, which can be applied to massive data and has very good scalability. Experimental results show that MPMF can not only achieve higher recommendation accuracy, but also improve the efficiency of the algorithm in sparse and unbalanced data sets compared with other related algorithms.
목차
1. Introduction
2. Related Work
3. The Definition of Fundamental Matrix Factorization Model
4. The Improved Probability Matrix Factorization Algorithm
4.1. The Traditional Recommendation Algorithm model-PMF
4.2. Improved Probability Matrix Factorization Algorithm—MPMF
5. Dataset and Metrics
5.1. Experiment Environment
5.2. Dataset
5.3. Metrics
6. Experimental Analysis
6.1. Experiment Scheme
6.2. The Impacts of Dimension D on Running Time of PMF
6.3. Comparison of RMSE in Training Set and Testing Set
6.4. Impacts of Dimension D on Recommendation Precision
6.5. Comparison of Recommendation Accuracy
6.6. Analysis of Time Complexity of MPMF
7. Conclusion
Acknowledgements
References