Behavior-Based Machine Learning Approaches for Malware Detection

The present research investigates the implementation of various methods of machine learning to analyze fluid information from executable files for the purpose identify contamination. We extracted feature vectors from a dataset that contained both malicious and benign samples. Standard metrics like F1-score, recall, and precision were used to train and evaluate a range of classifiers, including Decision Trees, Random Forests, Support Vector Machines, and Logistic Regression. Our test findings show that the models successfully distinguish between harmful and benign files, with Random Forest providing the highest level of consistency in all evaluation measures. These Results demonstrate the promise of machine learning methods to improve systems for detecting malware.