ProtoGate: Prototype-based Neural Networks with Global-to-local Feature Selection for Tabular Biomedical Data

TL;DR

ProtoGate is a neural network model designed for high-dimensional, low-sample-size biomedical tabular data, combining global and local feature selection with prototype-based prediction to improve accuracy and interpretability.

Contribution

ProtoGate introduces a novel prototype-based neural approach that balances global and local feature selection, addressing co-adaptation and enhancing performance on HDLSS data.

Findings

Outperforms state-of-the-art methods in prediction accuracy

Provides high-fidelity feature selection and interpretability

Effective on both synthetic and real-world datasets

Abstract

Tabular biomedical data poses challenges in machine learning because it is often high-dimensional and typically low-sample-size (HDLSS). Previous research has attempted to address these challenges via local feature selection, but existing approaches often fail to achieve optimal performance due to their limitation in identifying globally important features and their susceptibility to the co-adaptation problem. In this paper, we propose ProtoGate, a prototype-based neural model for feature selection on HDLSS data. ProtoGate first selects instance-wise features via adaptively balancing global and local feature selection. Furthermore, ProtoGate employs a non-parametric prototype-based prediction mechanism to tackle the co-adaptation problem, ensuring the feature selection results and predictions are consistent with underlying data clusters. We conduct comprehensive experiments to evaluate the performance and interpretability of ProtoGate on synthetic and real-world datasets. The results show that ProtoGate generally outperforms state-of-the-art methods in prediction accuracy by a clear margin while providing high-fidelity feature selection and explainable predictions. Code is available at https://github.com/SilenceX12138/ProtoGate.