A Projected Subgradient Method for Scalable Multi-Task Learning
dc.contributor.advisor | Trevor Darrell | en_US |
dc.contributor.author | Quattoni, Ariadna | en_US |
dc.contributor.author | Carreras, Xavier | en_US |
dc.contributor.author | Collins, Michael | en_US |
dc.contributor.author | Darrell, Trevor | en_US |
dc.contributor.other | Vision | en_US |
dc.date.accessioned | 2008-07-24T20:00:14Z | |
dc.date.available | 2008-07-24T20:00:14Z | |
dc.date.issued | 2008-07-23 | en_US |
dc.identifier.other | MIT-CSAIL-TR-2008-045 | en_US |
dc.identifier.uri | http://hdl.handle.net/1721.1/41888 | |
dc.description.abstract | Recent approaches to multi-task learning have investigated the use of a variety of matrix norm regularization schemes for promoting feature sharing across tasks.In essence, these approaches aim at extending the l1 framework for sparse single task approximation to the multi-task setting. In this paper we focus on the computational complexity of training a jointly regularized model and propose an optimization algorithm whose complexity is linear with the number of training examples and O(n log n) with n being the number of parameters of the joint model. Our algorithm is based on setting jointly regularized loss minimization as a convex constrained optimization problem for which we develop an efficient projected gradient algorithm. The main contribution of this paper is the derivation of a gradient projection method with l1ââ constraints that can be performed efficiently and which has convergence rates. | en_US |
dc.format.extent | 8 p. | en_US |
dc.relation | Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory | en_US |
dc.relation | en_US | |
dc.title | A Projected Subgradient Method for Scalable Multi-Task Learning | en_US |