Prediction Focused Topic Models for Electronic Health Records

TL;DR

This paper introduces a prediction-focused topic model for EHR data that enhances interpretability of latent factors while maintaining prediction accuracy, outperforming existing models in coherence.

Contribution

The novel prediction-focused topic model selectively retains features that improve prediction, balancing interpretability and predictive performance in high-dimensional EHR data.

Findings

More coherent topics learned compared to existing methods

Maintains competitive prediction accuracy

Effective on EHR and movie review datasets

Abstract

Electronic Health Record (EHR) data can be represented as discrete counts over a high dimensional set of possible procedures, diagnoses, and medications. Supervised topic models present an attractive option for incorporating EHR data as features into a prediction problem: given a patient's record, we estimate a set of latent factors that are predictive of the response variable. However, existing methods for supervised topic modeling struggle to balance prediction quality and coherence of the latent factors. We introduce a novel approach, the prediction-focused topic model, that uses the supervisory signal to retain only features that improve, or do not hinder, prediction performance. By removing features with irrelevant signal, the topic model is able to learn task-relevant, interpretable topics. We demonstrate on a EHR dataset and a movie review dataset that compared to existing approaches, prediction-focused topic models are able to learn much more coherent topics while maintaining competitive predictions.