Mitigating Malware Threats on Emerging Technology: A Machine Learning Approach

Abstract

Malicious programs and malware threats lead to a substantial vulnerability and pose a fundamental problem. Nowadays, smart devices are becoming more common, and consequently, the risk of malware intrusion is highly observed. This paper presents a comprehensive exploration from the initial to the final phase of an effective strategy and the deployment of a model to detect malware efficiently. The proposed Mitigating Malware Threats on Emerging Technology framework “MMTET” will help mitigate the risk of intrusion. This study explores the complexity of handling datasets. Random Forests and Decision Trees serve as machine learning algorithms for training and testing. Starting with a data collection method to obtain relevant parameters, this paper highlights the importance of well-curated datasets in training using effective machine learning models. Data analysis follows a statistical approach, and the visualization tools are used for identifying inherent biases, imbalances, and trends in datasets. For boosting the dataset’s quality, feature engineering and selection take a central stage to balance the data with new methodologies and detect relevant features with correlation analysis. Experimental result shows that Random Forest has the best performance compared to other methods obtained from different algorithms, with accuracy 98.30%.