DEVELOPMENT OF ADAPTIVE ALGORITHMS FOR AUTOMATED FORMATION OF CONTROL DECISIONS IN INTELLIGENT SYSTEMS FOR MANAGING RESIDUAL KNOWLEDGE OF STAFF

Tarasov Vyacheslav Sergeevich

In the context of dynamic technological development and the imperative of continuous professional education (Life-Long Learning), increasing the effectiveness of in-house training (IFT) by optimizing the residual knowledge (RK) of employees is becoming critical. This research article is devoted to the development of scientific and methodological support for the automated formation of management decisions in organizational systems based on monitoring information on personnel performance. The intelligent core of the system is neuro-fuzzy networks of the Adaptive Resonance Theory (ART) family, in particular Fuzzy ARTMAP, the clustering results of which serve as the basis for generating targeted pedagogical and organizational interventions. A formalization of the problem of selecting the optimal control action is presented, and multi-level algorithms are developed that ensure the transition from cluster-oriented (group) recommendations to highly personalized interventions that take into account the detailed knowledge profile and contextual information of each employee. The developed approach enables the implementation of a closed-loop management of residual knowledge, significantly increasing the adaptability of educational trajectories and the overall competence of personnel.

Keywords:residual knowledge, continuous learning, Fuzzy ARTMAP, data clustering, intelligent decision support, control actions, personalization, organizational system.