METHOD FOR DETERMINING THE OPTIMAL HYPERPARAMETERS OF GRADIENT BOOSTING MODELS FOR CLASSIFYING THE PROCESSES OF FUNCTIONING OF ENERGY-DEPENDENT MEMORY OF THE ORGANIZATION'S INFORMATION SYSTEM ASSETS

Detecting anomalous activity in the energy-dependent memory of an organization's information system assets is an actual task to ensure their protection. Gradient Boosting methods [1], such as HistGradientBoosting, LightGBM, XGBoost, and AdaBoost, demonstrate high accuracy in classification tasks, but their performance significantly depends on the configuration of hyperparameters [2-5].