Abstract:

Modern cyber threats and the dynamic development of malware pose new demands on cybersecurity training, requiring a shift from traditional signature approaches to intelligent methods based on machine learning and neural networks. In this context, this article presents an educationally oriented model for training in malware detection that integrates theoretical knowledge with practical skills by working with real data and modern analytical tools.

The proposed approach is structured as a learning module, including sequential stages: introduction to the main types of malwares, analysis of PE files, feature extraction and building models with different neural network architectures. Particular emphasis is placed on active learning through practical exercises, in which students independently develop, train and evaluate classification models. This process supports the development of key competencies such as analytical thinking, interpretation of results and critical assessment of the effectiveness of algorithms.

The educational value of the proposed model is expressed in increasing student engagement, better understanding of real-world cybersecurity challenges, and building practical skills. The results of the approach show that students not only master the theoretical foundations, but also successfully apply their knowledge in practical scenarios. The model is applicable in both face-to-face and distance learning and can be integrated into university courses to prepare highly qualified specialists capable of dealing with real-world cyber threats.

Keywords:

Malware detection, machine learning, neural networks, PE file analysis, practical learning.