აბსტრაქტი. The use of smartphones has wisely evolved in the 20th century. Many people all over the world can connect to their smartphones in a variety of ways. Some invaders are leveraging the power of the rapidly growing smartphone usage to Developing rogue Android applications to steal handsets' sensitive data To address these grave issues, a malicious program that is both effective and efficient is required. Numerous malware detection programs have historically been built, but some of them are not able to identify recently developed malware programs or programs contaminated with different Trojan horses, worms, and spyware. The software for detecting fraudulent programs can be enhanced especially thanks to ML algorithms. The system uses ML classification algorithms like Support Vector Machine (SVM) to improve the malware application detection in the proposed system. The proposed ML classification and fusion algorithms will improve performance metrics like the accuracy of malware application exposure and decrease the complexity of the detection process. The suggested approach integrates detecting software with a training application that users can install on Android cellphones to flag hazardous activities when they are accessible.