Розробка комплексного показника діагностики масованих ракетних ударів

Objective. Enhancing the efficiency of diagnosing the threat level of massive missile strikes by developing a comprehensive indicator. Methodology. The study examines the process of developing a comprehensive indicator based on a dataset of massive missile strikes. This involves preliminary data processing, the development of a comprehensive indicator model, and the integration of individual indicators. One of the integrated indicators is assigned a weight coefficient, which is determined using artificial intelligence methods and constrained by a sigmoid activation function. A comparative analysis of the proposed comprehensive indicator against existing indicators was conducted based on the standard deviation criterion. The assessments obtained using the comprehensive indicator are employed to determine the threat level of massive missile strikes. Results. Based on an existing dataset of massive missile strikes on Ukraine, a comprehensive indicator has been developed, consolidating attack characteristics into a unified assessment. The comprehensive indicator's evaluations regarding massive missile strikes are utilized to determine the threat level (Cluster 1 – low threat level, Cluster 2 – high threat level). Scientific novelty. The proposed comprehensive indicator model differs from existing approaches in that its integrated indicators account for mean values and variations in assessments, serving as a prototype of the regularization concept. As a result, the standard deviation is reduced to 0.0925, whereas the existing approach demonstrates a deviation of 0.447 on a single experimental set of assessments. Practical significance. The proposed comprehensive indicator of massive missile strikes serves as an additional measure for determining the state's threat level or may be considered an element of a decision-making system.