Development of a load classifier for a model for assessing the state of a cyberphysical system operator

Penin Andrei Semenovich

During the research work, the main biological markers of the human body were studied and those of them that met the requirements were selected for further research. A classifier of the employee's workload level was developed based on the decision tree method, the classification accuracy was 76%, the value of the loss function was 0.23. In the future, the classifier of the employee's workload level and biological markers will be combined into a model for assessing the state of the operator of a cyber-physical system.

Keywords:neural networks, multi-class classification, biomarkers, models, systems, state.