Conditional Computation in Neural Networks for faster models

• DOI: 10.48550/arxiv.1511.06297
• Publication Date: 2015-01-01
• Authors: Bengio, Emmanuel; Bacon, Pierre-Luc; Pineau, Joelle; Precup, Doina
• Keywords: FOS: Computer and information sciences; Computer Science - Machine Learning; 01 natural sciences; 0105 earth and related environmental sciences; Machine Learning (cs.LG)

Deep learning has become the state-of-art tool in many applications, but the evaluation and training of deep models can be time-consuming and computationally expensive. The conditional computation approach has been proposed to tackle this problem (Bengio et al., 2013; Davis & Arel, 2013). It operates by selectively activating only parts of the network at a time. In this paper, we use reinforcement learning as a tool to optimize conditional computation policies. More specifically, we cast the problem of learning activation-dependent policies for dropping out blocks of units as a reinforcement learning problem. We propose a learning scheme motivated by computation speed, capturing the idea of wanting to have parsimonious activations while maintaining prediction accuracy. We apply a policy gradient algorithm for learning policies that optimize this loss function and propose a regularization mechanism that encourages diversification of the dropout policy. We present encouraging empirical results showing that this approach improves the speed of computation without impacting the quality of the approximation.<br/>ICLR 2016 submission, revised<br/>