Using Neural Architecture Search for Improving Software Flaw Detection in Multimodal Deep Learning Models

TL;DR

This paper shows that combining neural architecture search with multimodal deep learning models significantly improves software flaw detection performance on benchmark datasets.

Contribution

It introduces a NAS framework adapted for software flaw detection, enhancing existing multimodal models' effectiveness.

Findings

Improved detection accuracy on Juliet Test Suite

Effective adaptation of NAS for software flaw detection

Demonstrated superiority over baseline models

Abstract

Software flaw detection using multimodal deep learning models has been demonstrated as a very competitive approach on benchmark problems. In this work, we demonstrate that even better performance can be achieved using neural architecture search (NAS) combined with multimodal learning models. We adapt a NAS framework aimed at investigating image classification to the problem of software flaw detection and demonstrate improved results on the Juliet Test Suite, a popular benchmarking data set for measuring performance of machine learning models in this problem domain.