Kishan Wimalawarne




  • We define the Scaled Latent Trace Norm that can be effciently used for regularising tensors with flat structures and high variations in multilinear rank. We apply the scaled latent trace norm to multilinear multi-task learning prolems and it out perform other tensor norms.
  • MATLAB implementation:













School dataset



  • Reference

Wimalawarne, K., Sugiyama, M., & Tomioka, R. Multitask learning meets tensor factorization: Task imputation via convex optimization. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems 27, pp.2825-2833, 2014. (Presented at Neural Information Processing Systems (NIPS2014), Montreal, Quebec, Canada, Dec. 8-11, 2014)