Learning Multiple Tasks with Multilinear Relationship Networks

Published in Neural Information Processing Systems (NIPS), 2017, 2017

Mingsheng Long, Zhangjie Cao, Jianmin Wang, Philip S. Yu. Neural Information Processing Systems NIPS 2017.

[Conference Version] [Arxiv] [Code]


Deep networks trained on large-scale data can learn transferable features to promote learning multiple tasks. Since deep features eventually transition from general to specific along deep networks, a fundamental problem of multi-task learning is how to exploit the task relatedness underlying parameter tensors and improve feature transferability in the multiple task-specific layers. This paper presents Multilinear Relationship Networks (MRN) that discover the task relationships based on novel tensor normal priors over parameter tensors of multiple task-specific layers in deep convolutional networks. By jointly learning transferable features and multilinear relationships of tasks and features, MRN is able to alleviate the dilemma of negative- transfer in the feature layers and under-transfer in the classifier layer. Experiments show that MRN yields state-of-the-art results on three multi-task learning datasets.