Year: 2025
Type: conference-paper
Abstract: This paper presents an example of a neuro-fuzzy system model designed for time series analysis in the context of information security, with a focus on anomaly detection in network traffic. The main goal of the model is to identify anomalies, which helps prevent cyberattacks and unauthorized access. The model consists of two key components: an artificial neural network (ANN) for processing input data and fuzzy logic for decision-making based on the ANN’s outputs, accounting for data uncertainty. The paper outlines the stages of data preprocessing, training the neural network on historical data, forming fuzzy rules based on the ANN’s outputs, and final defuzzification, which leads to obtaining a clear assessment of the threat level. The provided example illustrates how the combination of ANN and fuzzy logic can effectively analyze time series and adapt to changes in system behavior, significantly reducing the risk of cyberattacks.
Authors: R. Doroshenko, A. Prokofyev, K. Starodubov
Institution: MIREA - Russian Technological University
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4409416448