Integrating Comparison of Malware Detection Classification using LGBM and XGB Machine Learning Algorithms

Malware detection is a critical component of computer system security. Additionally Signature-based approaches are futile in detection of zero-day attacks and polymorphic infections. As a result, machine learning-based detection becomes necessary. The goal of this study was to find the optimum feature extraction, feature representation, and classification algorithms for using on top of LGBM (Light GBM) and XGB (Extreme Gradient Boosting) classifiers to get the best accuracy. The dataset for this investigation included 1156 malware files from nine different families, as well as 984 benign files in various formats. This paper outlines recommended approaches for malware categorization and detection using machine learning, as well as implementation guidance. Furthermore, the research conducted can serve as a future effort in the meadow of malware analysis using machine learning methodologies will be built on this basis. To increase the feature set and lower the detection false positive rate, this research employs the LGBM (Light GBM) and XGB (Extreme Gradient Boosting) classifiers. On a benchmark classification, we used a range of machine learning methods in our tests. Proposed system has a malware detection rate comparable to the state of the art and shows promise in categorization forecasts, allowing it to deal with obfuscation and the growing volume of malware.