MALWARE CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK TO IMPROVE INDONESIAN GOVERNMENT CYBERSECURITY

As the dependence on information technology and cyberspace intensifies across various critical sectors, it also presents a potential medium for cyber attacks. Particularly, the rapidly evolving nature of malware, with its myriad variants displaying distinct characteristics, often leaves victims grappling with effective mitigation. Current malware detection technologies, although numerous, are often deemed insufficient due to their inability to provide robust classification, an essential aspect that facilitates efficient cybersecurity analysis. This research addresses this pressing issue by harnessing Convolutional Neural Networks (CNNs) Inception-V3 for the classification of malware, with an aim to enhance the speed and effectiveness of malware mitigation efforts, especially in Indonesian Government. CNNs, subtypes of artificial neural networks, are mainly used for visual data examination. They employ a mathematical operation known as convolution in one or more of their layers, making them particularly adept at handling pixel data for tasks such as image recognition, processing, and classification. This research serves as a significant stride towards to improve Indonesian Government Cybersecurity on dealing with the cyber threat and reducing dependence on the use of signature-based malware detection which is considered to have many weaknesses.

Keywords: Malware Classification, Convolutional Neural Network