Ensemble Learning for Malware Classification: A Performance Evaluation Using Random Forest and Hist Gradient Boosting

As technology advances, malware has become a major threat to information systems and software security. The increasing sophistication of malware requires fast and accurate detection methods to mitigate potential cyber threats. Traditional detection techniques often struggle to adapt to new malware variants, making machine learning a promising solution. This study applies ensemble learning for malware classification, utilizing Random Forest and Hist Gradient Boosting algorithms. The research is conducted on a dataset of Portable Executable (PE) files, commonly used in Windows malware. The model's performance is evaluated based on accuracy, precision, recall, and F1-score to ensure reliable detection. Experimental results demonstrate excellent classification performance, achieving 99.24% accuracy, 99.15% precision, 99.83% recall, and 99.49% F1-score. These findings indicate that ensemble learning is highly effective for distinguishing malware from benign files, making it a valuable tool in cybersecurity. However, this study does not incorporate hyperparameter tuning, which could further optimize performance. Future research should explore fine-tuning techniques and test on more diverse datasets to improve generalizability. The results highlight ensemble learning’s potential in malware detection, contributing to the development of more robust and adaptive cybersecurity solutions against evolving threats.