Clustering by Nonnegative Matrix Factorization Using Graph Random Walk (NMFR)

• nmfr_auto.m for small-scale data, automatically selecting the random walk extent
• nmfr.m for large-scale data, (random walk extent needed, suggested value 0.8)

Please give reference to our contribution below if you use our NMFR codes in your publishable work.

• Zhirong Yang, Tele Hao, Onur Dikmen, Xi Chen, Erkki Oja. Clustering by Nonnegative Matrix Factorization Using Graph Random Walk. In Advances in Neural Information Processing Systems 25 (NIPS 2012), Lake Tahoe, USA. [pdf] [supplemental document] [clustering experiment scripts and datasets]