Dynamic feature selection in medical predictive monitoring by reinforcement learning

TL;DR

This paper introduces a reinforcement learning-based method for dynamic, patient-specific feature selection in clinical time-series prediction, improving accuracy under cost constraints.

Contribution

It presents a novel reinforcement learning framework for dynamic feature selection in time-series clinical data, capable of handling non-differentiable models and optimizing feature subsets over time.

Findings

Outperforms baseline feature selection methods in clinical tasks

Effective under strict cost limitations

Seamless integration with various prediction models

Abstract

In this paper, we investigate dynamic feature selection within multivariate time-series scenario, a common occurrence in clinical prediction monitoring where each feature corresponds to a bio-test result. Many existing feature selection methods fall short in effectively leveraging time-series information, primarily because they are designed for static data. Our approach addresses this limitation by enabling the selection of time-varying feature subsets for each patient. Specifically, we employ reinforcement learning to optimize a policy under maximum cost restrictions. The prediction model is subsequently updated using synthetic data generated by trained policy. Our method can seamlessly integrate with non-differentiable prediction models. We conducted experiments on a sizable clinical dataset encompassing regression and classification tasks. The results demonstrate that our approach outperforms strong feature selection baselines, particularly when subjected to stringent cost limitations. Code will be released once paper is accepted.