Paper Title
A Hybrid Artificial Classifier for Android Mobile Malware Detection

Abstract
Penetrating Android security defense systems occurred more frequently in the recent years. These encouraging researchers to investigate and evaluate the execution behaviors of malicious applications in mobile systems, especially, in Android Operating Systems (OS).The main challenges of those detection systems is having a high rate of false alarm, which is related to pretending behaviors of malwares in communicating systems that mostly detected and classified as normal. This paper proposes mimicking a biological phenomenon named Co-stimulation, which occurred within the human immune system's activities. This phenomenon confirms that ID of an abnormal and normal entities depending on two different set of cell’s features. This confirmation avoids attacking self-cells, which means eliminating the false alarm. Same scenario has been simulated in mobile malware detection process using Artificial Neural Network (ANN). This work proposes a combination of different sets of mobile malware features to train detection systems. The hybrid set grantees increasing the accuracy rate of mobile malware detection and classification. Index Terms - ANN, Co-stimulation phenomenon, Android Security, Mobile Malwares.