Clustered Multi-Task Learning: A Convex Formulation
Abstract:In multi-task learning several related tasks are considered simultaneously, with the hope that by an appropriate sharing of information across tasks, each task may benefit from the others. In the context of learning linear functions for supervised classification or regression, this can be achieved by including a priori information about the weight vectors associated with the tasks, and how they are expected to be related to each other. In this paper, we assume that tasks are clustered into groups, which are unknown beforehand, and that tasks within a group have similar weight vectors. We design a new spectral norm that encodes this a priori assumption, without the prior knowledge of the partition of tasks into groups, resulting in a new convex optimization formulation for multi-task learning. We show in simulations on synthetic examples and on the IEDB MHC-I binding dataset, that our approach outperforms well-known convex methods for multi-task learning, as well as related non-convex methods dedicated to the same problem.
暂无分享,去 创建一个
[1] Steven A. Orszag,et al. CBMS-NSF REGIONAL CONFERENCE SERIES IN APPLIED MATHEMATICS , 1978 .
[2] G. Wahba. Spline models for observational data , 1990 .
[3] Tomaso A. Poggio,et al. Regularization Theory and Neural Networks Architectures , 1995, Neural Computation.
[4] R. Tibshirani. Regression Shrinkage and Selection via the Lasso , 1996 .
[5] Tom Heskes,et al. Task Clustering and Gating for Bayesian Multitask Learning , 2003, J. Mach. Learn. Res..
[6] Nello Cristianini,et al. Learning the Kernel Matrix with Semidefinite Programming , 2002, J. Mach. Learn. Res..
[7] Charles A. Micchelli,et al. Learning Multiple Tasks with Kernel Methods , 2005, J. Mach. Learn. Res..
[8] D. Heckerman,et al. Leveraging Information Across HLA Alleles/Supertypes Improves HLA-Specific Epitope Prediction , 2005 .
[9] Francis R. Bach,et al. Low-rank matrix factorization with attributes , 2006, ArXiv.
[10] Morten Nielsen,et al. A Community Resource Benchmarking Predictions of Peptide Binding to MHC-I Molecules , 2006, PLoS Comput. Biol..
[11] Massimiliano Pontil,et al. Multi-Task Feature Learning , 2006, NIPS.
[12] David Heckerman,et al. Leveraging Information Across HLA Alleles/Supertypes Improves Epitope Prediction , 2006, RECOMB.
[13] David Heckerman,et al. Leveraging Information Across HLA Alleles/Supertypes Improves Epitope Prediction , 2007, J. Comput. Biol..
[14] Joydeep Ghosh,et al. A framework for simultaneous co-clustering and learning from complex data , 2007, KDD '07.
[15] Jean-Philippe Vert,et al. Efficient peptide-MHC-I binding prediction for alleles with few known binders , 2008, Bioinform..