Web Based Malware & Virus Prediction System Using Machine Learning

Abstract

Malware-based cyber-attacks are increasing and have advanced in sophistication across practically all networks. Attacks using sophisticated malware are also the most complex, which makes it difficult to detect them. Advanced malware can hide many of its traces using various techniques, including metamorphic engines causing significant financial and privacy losses to a variety of organizations As a result, malware analysis methods now face a considerable barrier in predicting and detecting such threats. With the tremendous technological development and cutting-edge research, researchers and anti-virus organizations have begun using machine learning and deep learning methods for malware analysis and detection. Through this proposed work we have proposed a system to detect the malware on Windows systems. According to the types of features, appropriate algorithm is chosen to check the malware and virus. Total 4 different algorithms named LightGBM, XGboost, and Logistic Regression, CatBoost Regressor is used here. To handle the missing values, skewness check, correlation test, feature engineering is applied with accuracy of 74%. For practical feeling of the proposed system an E-commerce website also designed.