Multilinear multitask learning

Romera-Paredes, Bernardino, Aung, Min Hane, Bianchi-Berthouze, Nadia and Pontil, Massimiliano (2013) Multilinear multitask learning. In: 30th International Conference on Machine Learning, ICML 2013, 2013-06-16 - 2013-06-21.

Full text not available from this repository.

Abstract

Many real world datasets occur or can be arranged into multi-modal structures. With such datasets, the tasks to be learnt can be referenced by multiple indices. Current multitask learning frameworks are not designed to account for the preservation of this information. We propose the use of multilinear algebra as a natural way to model such a set of related tasks. We present two learning methods; one is an adapted convex relaxation method used in the context of tensor completion. The second method is based on the Tucker decomposition and on alternating minimization. Experiments on synthetic and real data indicate that the multilinear approaches provide a significant improvement over other multitask learning methods. Overall our second approach yields the best performance in all datasets.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: human-computer interaction,sociology and political science ,/dk/atira/pure/subjectarea/asjc/1700/1709
Related URLs:
Depositing User: LivePure Connector
Date Deposited: 27 Nov 2019 01:24
Last Modified: 22 Apr 2020 09:42
URI: https://ueaeprints.uea.ac.uk/id/eprint/73038
DOI:

Actions (login required)

View Item View Item