Sampling Through the Lens of Sequential Decision Making

TL;DR

This paper introduces a reinforcement learning-based adaptive sampling method called ASR, which dynamically optimizes sampling strategies during training to improve representation learning in tasks like information retrieval and clustering.

Contribution

It is the first to apply reinforcement learning to adaptively optimize sampling processes in representation learning, surpassing fixed or heuristic-based methods.

Findings

ASR outperforms traditional sampling methods in information retrieval.

ASR demonstrates superior clustering performance across datasets.

Identification of the 'ASR gravity well' phenomenon in sampling dynamics.

Abstract

Sampling is ubiquitous in machine learning methodologies. Due to the growth of large datasets and model complexity, we want to learn and adapt the sampling process while training a representation. Towards achieving this grand goal, a variety of sampling techniques have been proposed. However, most of them either use a fixed sampling scheme or adjust the sampling scheme based on simple heuristics. They cannot choose the best sample for model training in different stages. Inspired by "Think, Fast and Slow" (System 1 and System 2) in cognitive science, we propose a reward-guided sampling strategy called Adaptive Sample with Reward (ASR) to tackle this challenge. To the best of our knowledge, this is the first work utilizing reinforcement learning (RL) to address the sampling problem in representation learning. Our approach optimally adjusts the sampling process to achieve optimal performance. We explore geographical relationships among samples by distance-based sampling to maximize overall cumulative reward. We apply ASR to the long-standing sampling problems in similarity-based loss functions. Empirical results in information retrieval and clustering demonstrate ASR's superb performance across different datasets. We also discuss an engrossing phenomenon which we name as "ASR gravity well" in experiments.