Skip to content
Conference

A Data-Driven Hybrid CNN-LSTM Model for Software Defect Prediction

Aug 2026 · International Conference Computational Vision and Bio Inspired Computing · pp. 419-422 · 0 citations · 20 references

Abstract

Software Defect Prediction (SDP) is an important aspect of enhancing software quality by determining early on in the development lifecycle the modules that are likely to be error prone. Complex machine learning methods fail to provide hidden nonlinear relationships within complex software measures. Our study has suggested a hybrid deep learning system with convolutional neural networks (CNNs) and long short-term memory (LSTM) networks as a viable solution to software defect prediction to overcome this limitation. The data of 10,885 modules in 21 features were used to train the model on a real-world software measure dataset. The proposed CNN-LSTM architecture, following the preparation and optimization of the model, reached the test accuracy threshold of ${7 5. 1 5 \%}$, which further improved to ${8 1. 3 0 \%}$ with the addition of threshold optimization. These findings prove that the hybrid deep learning model is effective at capturing both local patterns and sequence dependencies in software metrics which substantially improves the prediction performance. The suggested solution offers an effective and dependable solution to software quality assurance and can assist developers in prioritizing testing and minimizing defects within large software systems.

View source

Similar papers

Open access 2026

FR-CNN: SOFTWARE FAULT PREDICTION USING A NOVEL HYBRID HIERARCHICAL MODEL

A novel hybrid model based on a fine-tuned fully connected fully connected deep neural network, a random forest, and a convolutional neural network that achieves competitive fault prediction performance while requiring lower computational complexity and training time than recent deep-learning approaches.

Mehrasa Jouybari, Alireza Tajary, M. Fateh et al. · 0 citations
Open access 2026

A Comparative Study of CNN and MLP for Software Defect Prediction Using PROMISE Datasets

This study applies both algorithms to detect software defects using 11 open-source datasets from the PROMISE repository and indicates that CNN outperforms MLP, achieving 86% prediction accuracy and an F1 score of 87.7%.

Ekhlas Tariq Hasan, S. Mohi-Aldeen · 0 citations
Open access Sep 2026

Comparative Evaluation of LSTM-Based Deep Learning Models for Software Effort Estimation

Results indicate that LSTM-GRU consistently delivers stable performance across datasets, while LSTM-Transformer achieves superior accuracy on complex datasets due to its self-attention mechanism.

Santa Margita, Eko Sediyono, S. Y. J. Prasetyo et al. · 0 citations
Open access Sep 2026

A Hybrid CNN–LSTM Deep Learning Framework for Automated Malware Classification using Spatial–Sequential Feature Fusion

The high rate of growth of advanced and highly obfuscated malware has made the use of conventional signature detection mechanisms less viable. In order to overcome this problem, this paper suggests a hybrid framework for malware classification based on the combination of Convolutional Neural Networks (CNNs) and Long Sh...

Chevella Anil Kumar, Saradha S, Mithila Ayyavoo et al. · 0 citations
Review Open access 2026

A Two-Stage Data-Driven Framework for App Quality Monitoring: Integrating CNN-BiLSTM and Laney p’ Control Charts

Digital public service applications generate large-scale user reviews that provide valuable feedback on service quality, but their unstructured nature makes continuous monitoring challenging. This study proposes a two-stage data-driven framework that integrates FastText-based Convolutional Neural Network-Bidirectional...

Novri Suhermi, M. Mashuri, Hollyviar Resnias Putri Zalukhu · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.