Learning to Branch for Multi-Task Learning

Autor:	Guo, Pengsheng, Lee, Chen-Yu, Ulbricht, Daniel
Rok vydání:	2020
Předmět:	Computer Science - Machine Learning Computer Science - Computer Vision and Pattern Recognition Statistics - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Training multiple tasks jointly in one deep network yields reduced latency during inference and better performance over the single-task counterpart by sharing certain layers of a network. However, over-sharing a network could erroneously enforce over-generalization, causing negative knowledge transfer across tasks. Prior works rely on human intuition or pre-computed task relatedness scores for ad hoc branching structures. They provide sub-optimal end results and often require huge efforts for the trial-and-error process. In this work, we present an automated multi-task learning algorithm that learns where to share or branch within a network, designing an effective network topology that is directly optimized for multiple objectives across tasks. Specifically, we propose a novel tree-structured design space that casts a tree branching operation as a gumbel-softmax sampling procedure. This enables differentiable network splitting that is end-to-end trainable. We validate the proposed method on controlled synthetic data, CelebA, and Taskonomy. Comment: Accepted at ICML 2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2006.01895 Zobrazit plný text záznamu View this record from Arxiv