Review Paper on Android Malware Detection based on the Dynamic Based Approach using Machine Learning

Review Paper on Android Malware Detection based on the Dynamic Based Approach using Machine Learning

Abhilasha Agrahari¹, Prof. Brajesh Patel²

¹M.Tech Scholar  (CTA),² Department of Computer Science & Engineering

Shri Ram Institute of Technology, Jabalpur, MP.

Abstract

This review paper presents a machine learning approach for detecting Android malware based on the co-existence of features. The proposed approach extracts feature from both benign and malicious apps and trains a machine learning model using co-existence features that have different distributions or patterns. This approach achieves a high detection rate of 99.5% and a low false positive rate of 0.5% on a large dataset of Android apps. The results demonstrate the effectiveness of the approach for handling unknown and zero-day malware variants. The proposed approach shows potential for practical applications in improving mobile security solutions.

Introduction:

Android malware detection is an important problem due to the rapid growth and diversity of mobile applications. Traditional methods, such as signature-based and rule-based, are ineffective against new and unknown malware variants. Machine learning approaches have been widely used for Android malware detection, and in this paper, we review a novel machine learning approach based on the co-existence of features for Android malware detection.