COMPARATIVE STUDY OF FEATURE-LEVEL AND DECISION-LEVEL FUSION STRATEGIES IN NEURAL NETWORK MODELS FOR MULTIMODAL PSYCHODIAGNOSTICS

This paper examines the effectiveness of feature-level and decision-level fusion strategies in neural network models for multimodal psychodiagnostics. Findings indicate that feature-level fusion enhances information extraction, while decision-level fusion improves diagnostic accuracy. A hybrid approach ensures greater reliability and expands applications in clinical practice