A Lightweight Neural Architecture Search Model for Medical Image Classification

TL;DR

This paper introduces ZO-DARTS+, a differentiable neural architecture search method that enhances efficiency and maintains high accuracy in medical image classification, reducing search time significantly.

Contribution

The paper proposes ZO-DARTS+, a novel NAS algorithm with sparse probability generation via bi-level optimization, improving search efficiency for medical image classification.

Findings

Matches state-of-the-art accuracy on five datasets

Reduces search time by up to three times

Demonstrates effectiveness across multiple medical imaging tasks

Abstract

Accurate classification of medical images is essential for modern diagnostics. Deep learning advancements led clinicians to increasingly use sophisticated models to make faster and more accurate decisions, sometimes replacing human judgment. However, model development is costly and repetitive. Neural Architecture Search (NAS) provides solutions by automating the design of deep learning architectures. This paper presents ZO-DARTS+, a differentiable NAS algorithm that improves search efficiency through a novel method of generating sparse probabilities by bi-level optimization. Experiments on five public medical datasets show that ZO-DARTS+ matches the accuracy of state-of-the-art solutions while reducing search times by up to three times.