Learning in Gated Neural Networks

Abstract

Gating is a key feature in modern neural networks including LSTMs, GRUs and sparsely-gated deep neural networks. The backbone of such gated networks is a mixture-of-experts layer, where several experts make regression decisions and gating controls how to weigh the decisions in an input-dependent manner. Despite having such a prominent role in both modern and classical machine learning, very little is understood about parameter recovery of mixture-of-experts since gradient descent and EM algorithms are known to be stuck in local optima in such models. In this paper, we perform a careful analysis of the optimization landscape and show that with appropriately designed loss functions, gradient descent can indeed learn the parameters accurately. A key idea underpinning our results is the design of two {\em distinct} loss functions, one for recovering the expert parameters and another for recovering the gating parameters. We demonstrate the first sample complexity results for parameter recovery in this model for any algorithm and demonstrate significant performance gains over standard loss functions in numerical experiments.

References in corpus (14)

• Sequence to Sequence Learning with Neural Networks
• On the Properties of Neural Machine Translation: Encoder-Decoder Approaches
• DRAW: A Recurrent Neural Network For Image Generation
• Global Optimality of Local Search for Low Rank Matrix Recovery
• Gradient Descent Learns Linear Dynamical Systems
• Learning Factored Representations in a Deep Mixture of Experts
• Recovery Guarantees for One-hidden-layer Neural Networks
• On the Convergence Rate of Training Recurrent Neural Networks
• On the low-rank approach for semidefinite programs arising in synchronization and community detection
• On the Computational Efficiency of Training Neural Networks
• Towards Binary-Valued Gates for Robust LSTM Training
• Learning One-hidden-layer Neural Networks under General Input Distributions
• Convergence Rates for Gaussian Mixtures of Experts
• Robust Spectral Filtering and Anomaly Detection