On the Improvement of Eye Tracking-Based Cognitive Workload Estimation Using Aggregation Functions

Area of Science:

• Human Factors
• Artificial Intelligence
• Cognitive Science

Background:

• Cognitive workload quantifies mental effort and is crucial for monitoring mental fatigue.
• Accurate cognitive workload estimation is vital for high-responsibility professions and educational material design.
• Existing methods for cognitive workload assessment can be improved through advanced computational techniques.

Purpose of the Study:

• To investigate the influence of diverse fuzzy and non-fuzzy aggregation functions on cognitive workload estimation quality.
• To evaluate the effectiveness of machine learning models in conjunction with aggregation functions for workload assessment.
• To explore the potential of aggregation processes for enhancing classification accuracy in cognitive workload recognition.

Main Methods:

• Application of various classic machine learning models to cognitive workload estimation.
• Systematic examination of over 2000 aggregation operators, including fuzzy and non-fuzzy types.
• In-depth experimental analysis to assess the performance of different aggregation functions.

Main Results:

• The study demonstrated the applicability of aggregation functions for cognitive workload estimation.
• Machine learning models combined with aggregation functions significantly improved classification results.
• Extensive experiments confirmed the effectiveness of the aggregation-based approach.

Conclusions:

• The integration of classical machine learning models with aggregation methods offers a high-quality solution for cognitive workload recognition.
• This approach achieves high accuracy while maintaining low computational costs.
• The findings support the use of aggregation functions as a valuable tool for improving mental fatigue monitoring and educational task design.