Nowadays, malware attacks have grown more dynamic and complicated. However, artificial intelligence approaches have also emerged as the major aspects of cybersecurity as these approaches are better equipped to handle contemporary malware threats. The article deals with the design of various feature selection techniques and ensemble approaches for android malware detection. The authors propose a new approach to malware detection that relies on supervised learning techniques using different feature selection algorithms. In this paper, the methodology relies over the selection of the combination of machine learning and feature selection techniques (FeatBoost, SFS, RFE) to produce the best possible solution for the detection of android malwares. The methodology is applicable to worldwide cyber security systems because attacks by malware in Africa (let’s say) are analogous to attacks globally, thus the authors developed the algorithm to help resolve an international issue. The experimental analysis revealed that the extra trees classifier achieved the highest accuracy of 99.55% when applying the RFE feature selection technique using the estimator as XGBoost and concludes that RFE is a better feature selection technique than SFS and FeatBoost.