EEG Sinyalleri Kullanılarak Zihinsel İş Yükü Seviyelerinin Sınıflandırılması Classification of Mental Workload Levels by Using EEG Signals

AKMAN AYDIN, Eda · 2021 · Crossref

DOI: 10.2339/politeknik.794655

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the objective classification of mental workload levels using electroencephalogram (EEG) signals. Mental workload, defined as the cognitive capacity required to perform tasks, is critical for evaluating performance in high-stakes environments such as aviation, traffic control, and surgery. Traditional assessment methods rely on subjective self-reports, which are limited by individual expression abilities. To overcome this, the research proposes an automated framework to classify mental workload into low, moderate, and high levels using fractal dimension algorithms for feature extraction and machine learning classifiers. The methodology utilizes the Simultaneous Task EEG Workload (STEW) dataset, comprising EEG recordings from 48 healthy male participants performing a multitasking test. EEG signals were recorded via 14 electrodes at 128 Hz. The study employs two fractal dimension algorithms for feature extraction: Katz Fractal Dimension (KFD) and Higuchi Fractal Dimension (HFD). These features are fed into a multiclass classification framework using Error Correcting Output Coding (ECOC) with a one-vs-all approach. Three classifiers are evaluated within this framework: Support Vector Machines (SVM), k-Nearest Neighbors (k-NN), and Quadratic Discriminant Analysis (QDA). The system’s performance is assessed using 10-fold cross-validation, measuring accuracy, sensitivity, and Cohen’s Kappa coefficient. The results demonstrate that the combination of HFD features with SVM-ECOC yields the highest classification performance. Specifically, the HFD-SVM-ECOC model achieved a classification accuracy of 95.39% and a Cohen’s Kappa value of 0.89. In comparison, the KFD-SVM-ECOC model achieved 78.74% accuracy and a Kappa of 0.52. Other classifier combinations performed lower; for instance, HFD with k-NN reached 92.80% accuracy, while HFD with QDA reached 73.76%. Sensitivity analysis revealed that the model was most effective at distinguishing low workload levels, while moderate workload levels presented the greatest classification challenge across all methods. The study concludes that HFD combined with SVM-ECOC is a highly effective method for the multiclass classification of mental workload. The findings highlight the superiority of HFD over KFD in capturing the complexity of EEG signals related to cognitive load. This approach offers a robust, objective tool for real-time monitoring of mental workload, with significant implications for human-computer interaction, adaptive systems, and safety-critical operational environments.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• neuro workload indices
• drowsiness detection algorithms
• workload measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data