Hybrid deep learning for mental workload classification using EEG with enhanced preprocessing and interpretability
Osama Abdelrahman, Chew XinYing, Esraa Faisal Malik, Khaw Khai Wah, Cheong Zhi Lin, Teoh Wei Lin · 2026
Read Hybrid deep learning for mental workload classification using EEG with enhanced preprocessing and interpretability by Osama Abdelrahman, Chew XinYing, Esraa Faisal Malik, Khaw Khai Wah, Cheong Zhi Lin, Teoh Wei Lin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0352882.txt.
Description
Osama Abdelrahman and five colleagues classify mental workload from brainwaves in this 2026 study. Knowing when an operator is overloaded matters in cockpits and operating theatres. Brainwave systems promise that knowledge, but noise and individual differences defeat many models. Their hybrid framework attacks all three weaknesses at once. An autoencoder cleans and compresses the signals. An attention module then spotlights the most informative channels and places. A memory network reads how the patterns unfold over time.
Tested with each person held out in turn, the system averaged about eighty-four percent accuracy. It sorted four workload levels and beat the baseline models overall. Removing any single component hurt performance, confirming each part earns its place. Tuning settled on ten-second windows as the sweet spot. Because the attention module can show what it focused on, the system explains itself better than most black boxes, a real virtue in safety-critical monitoring.
Catalog record
| CREATOR | Osama Abdelrahman, Chew XinYing, Esraa Faisal Malik, Khaw Khai Wah, Cheong Zhi Lin, Teoh Wei Lin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54464 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |