Utilizing Semantic Textual Similarity for Clinical Survey Data Feature Selection

TL;DR

This paper explores using semantic textual similarity scores derived from language models to improve feature selection in clinical survey data, leading to better predictive model performance.

Contribution

It introduces a novel approach of leveraging feature name semantics via language models for feature selection in clinical survey datasets.

Findings

STS-based feature selection outperforms traditional methods

Improved model performance with semantic feature selection

Effective in high-dimensional, low-sample clinical data

Abstract

Survey data can contain a high number of features while having a comparatively low quantity of examples. Machine learning models that attempt to predict outcomes from survey data under these conditions can overfit and result in poor generalizability. One remedy to this issue is feature selection, which attempts to select an optimal subset of features to learn upon. A relatively unexplored source of information in the feature selection process is the usage of textual names of features, which may be semantically indicative of which features are relevant to a target outcome. The relationships between feature names and target names can be evaluated using language models (LMs) to produce semantic textual similarity (STS) scores, which can then be used to select features. We examine the performance using STS to select features directly and in the minimal-redundancy-maximal-relevance (mRMR) algorithm. The performance of STS as a feature selection metric is evaluated against preliminary survey data collected as a part of a clinical study on persistent post-surgical pain (PPSP). The results suggest that features selected with STS can result in higher performance models compared to traditional feature selection algorithms.