Cognitive neural network for data mining

Cognitive neural network-based software system for data mining and processing. The system implements a multi-layered cognitive learning architecture with attentional mechanisms based on differential equations and splines. Includes data preprocessing, decision explainability, and hyperparameter optimization modules. Supports working with textual, tabular and graphical data.Implemented in Python using PyTorch, achieved classification accuracy of more than 95% on standard benchmarks.

Manager / participants: professor Bychkov O.S.