Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles

TL;DR

This paper introduces a stochastic gradient descent method for training diverse deep ensembles to generate multiple high-quality hypotheses, improving oracle error and interpretability across various tasks.

Contribution

It presents a novel, architecture-agnostic, parameter-free approach for producing diverse outputs in deep ensembles using stochastic gradient descent.

Findings

Lower oracle error compared to existing methods

Produces interpretable representations of task ambiguity

Effective across various tasks and deep architectures

Abstract

Many practical perception systems exist within larger processes that include interactions with users or additional components capable of evaluating the quality of predicted solutions. In these contexts, it is beneficial to provide these oracle mechanisms with multiple highly likely hypotheses rather than a single prediction. In this work, we pose the task of producing multiple outputs as a learning problem over an ensemble of deep networks -- introducing a novel stochastic gradient descent based approach to minimize the loss with respect to an oracle. Our method is simple to implement, agnostic to both architecture and loss function, and parameter-free. Our approach achieves lower oracle error compared to existing methods on a wide range of tasks and deep architectures. We also show qualitatively that the diverse solutions produced often provide interpretable representations of task ambiguity.