I am working on a project for collaborative filtering (CF), i.e. completing a partially observed matrix or more generally tensor. I am a newbie to the field, and for this project eventually I have to compare our method to other well-known ones that nowadays, proposed methods are compared against them, namely state-of-the-art in CF.
My search revealed the following methods. Indeed I came across them by looking at some of these papers and their references, or by looking at experiments section when they do comparisons. I would be happy to know for a new proposed method and to do a comparison with SoTA, which of the following would be a good pick to do so? If not among them, I would be happy to know a good representative.
Based on Matrix Factorization:
- Weighted Low Rank Approximation (ICML 2003)
- Modeling User Rating Profiles For Collaborative Filtering (NIPS 2003)
- The Multiple Multiplicative Factor Model For Collaborative Filtering (ICML 2004)
- Fast Maximum Margin Matrix Factorization for Collaborative Prediction (ICML 2005)
- Probabilistic Matrix Factorization (NIPS 2007)
- Bayesian Probabilistic Matrix Factorization (ICML 2008)
- Regression-based Latent Factor Models (KDD 2009)
- Non-linear Matrix Factorization with Gaussian Processes (ICML 2009)
- Dynamic Poission Factorization (ACM Conference on Recommender Systems 2015)
Based on Tensor Factorization:
- Incorporating Contextual Information in Recommender Systems Using a Multidimensional Approach (ACM Transactions on Information Systems (TOIS) 2005)
- Bayesian Probabilistic Tensor Factorization (SIAM Data Mining 2010)
- Low-rank tensor completion by Riemannian optimization (BIT Numerical Mathematics 54.2 (2014))