Development and implementation of an attack detection system based on neural networks to protect information of the federal budget institution Administration of the Amur Basin of Inland Waterways

In the modern information environment, characterized by an increase in the volume and value of processed information, ensuring information security is of paramount importance. Traditional attack detection methods often prove ineffective in the face of complex cyber threats. The article proposes an approach to improving the level of information security of the Federal Budget Institution «Administration of the Amur Basin of Inland Waterways» by developing an attack detection system based on the use of artificial neural networks capable of adaptive learning and pattern recognition. An artificial neural network is a system of interconnected neurons capable of processing large amounts of data and identifying hidden patterns. To solve the problem of detecting attacks, a deep learning algorithm has been chosen, which has proven its effectiveness in solving image recognition problems. The simulation results show that a neural network trained on a specially generated database is able to classify e-mail messages as spam or «regular», which helps optimize mail handling and improves information security of the Federal Budget Institution Administration of the Amur Basin of Inland Waterways. The proposed solution represents an effective approach to improving the information security of the Federal Budget Institution Amur Basin Administration of Inland Waterways by timely detecting and blocking spam messages that may contain malicious software or phishing links.