"APPLICATION OF INTELLIGENT DATA ANALYSIS APPROACHES FOR PRELIMINARY CYBER‑THREAT ASSESSMENT OF CRITICAL INFRASTRUCTURE FACILITIES"

Based on current requirements for time‑series analysis in cyber‑threat detection, this paper examines forecasting methods for identifying anomalous requests to critical‑infrastructure assets. Special attention is paid to building models that account for both the current traffic dynamics and its rate of change. The approaches analyzed include AutoRegressive (AR), AutoRegressive Moving Average (ARMA), AutoRegressive Integrated Moving Average (ARIMA), and a sequential model with linear layer stacking. To ensure forecasting accuracy, we propose testing data stationarity and applying decomposition techniques to isolate seasonal components. The application of these methods enables the detection of anomalies indicative of potential cyber threats. The results demonstrate the feasibility of developing adaptive cyber‑defense systems capable of timely response to threats, minimizing the risk of cyber incidents while maintaining modeling precision.