An Effective Machine Learning and Deep Learning Algorithms for Android Malware Classification and Detection

With the growing reliance on mobile devices, malware attacks are becoming more frequent, particularly on Android phones, which hold 72.2% of the market share. Cybercriminals target smart phones using various tactics, including credential theft, surveillance, and malicious advertising. This paper presents a systematic review of machine learning-based techniques for detecting Android malware. Malware classification involves grouping malware into families based on their unique signature. This dataset is collected from kaggle repository. One of the most significant challenges facing mobile users today is malware. This study focuses on classifying emerging malware based on shared features of similar malware. It introduces a novel framework that categorizes malware samples into families and detects new malware for further analysis. To achieve this, Artificial Neural Network (ANN) technique is utilized. The proposed approach aids in identifying and eliminating new malware while effectively classifying them into their respective families. The Proposed ANN techniques are obtained the accuracy of 98%, which is 4% higher than the RF, 2% higher than the J48, and 3%, 1% higher than the SVM, and Navie Bayes.