How to do Cross-Validation for Deep Learning models on EEG Signals - Mental Workload Classification

Описание: This poster presents our work published at Neuroergonomics Conference 2021 (NEC2021) with the title “On Time-Series Cross-Validation for Mental Workload Classification from EEG Signals”.

To know more about the NEC2021 conference, please visit https://www.neuroergonomicsconference....

-----------------------------

In this work, we tackle the challenging issue of evaluating Deep Learning models for Mental Workload classification using Electroencephalogram (EEG) signals. In particular, we raise awareness of the issues involved in using traditional cross-validation techniques for time-series based data, such as EEG. We then propose a more suited Cross-Validation technique tailored to such data. We finally investigate how such a technique can be implemented for Mental Workload classification.

-----------------------------

This work was conducted in NeuraSearch Laboratory by Kunjira Kingphai and Dr Yashar Moshfeghi.

To know more about Kunjira Kingphai, please visit   / kunjira-kingphai-448832206  .

To know more about Dr Yashar Moshfeghi, please visit http://academic.yashmosh.com or   / yashar-moshfeghi  .

To know more about our research activities at NeuraSearch Laboratory, please follow us on Twitter (@NeuraSearch) and to get notified of future uploads please subscribe to our channel!

Thank you.

-----------------------------