Aplicação de redes autocodificadoras para a detecção de malware

Abstract

In recent years, devices such as smartphones and Internet of Things (IoT) equipment are becoming increasingly prevalent, to the point where they constitute the majority of devices connected to the Internet. Despite the differences in processing power, storage, and tasks made, these devices could deal with user-sensible data, and like any device, they can have vulnerabilities, making them attractive targets for malicious attacks. Overthe years, consumers have become increasingly concerned about the potential loss of privacy resulting from the use of these devices, which increases the need for security tools to protect these areas. Thus, the proposal of creating a malware detection model to analyze datasets from those devices emerges, which, unlike the solutions presented in other works, aims to apply the concepts of deep neural networks using the autoencoder algorithm.