Classifying the Stickiness of Mind-Wandering

Vries, Stijn de (2021) Classifying the Stickiness of Mind-Wandering. Bachelor's Thesis, Artificial Intelligence.

Preview	Text AI_BA_2021_StijnDeVries.pdf Download (579kB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (125kB)

Abstract

The experience of your mind wandering away from the task at hand is familiar to many. Sometimes disengaging from mind-wandering and returning to the task at hand can be very difficult. This can be disruptive and potentially dangerous if caution is required. Detection of this ”sticky” mind-wandering can be beneficial for safety and productivity reasons. The occurrence of sticky mind-wandering was predicted on the basis of electroencephalography (EEG) data recorded during a sustained attention to response task and a visual search task. The types of mind-wandering were assessed by means of questions inserted in the tasks that asked participants about their mental state, on or off task, as well as the content of their thoughts. A Logistic Regression (LR), Random Forest Classifier (RFC) and Support Vector Machine (SVM) were then trained on feature vectors obtained from temporal and spatial sampling points where a significant difference was observed between sticky and non-sticky mind-wandering. It was found that SVM (accuracy 62.2%) outperformed the LR (57%) and the RFC (58.1%), potentially because it is better able to deal with the high complexity of EEG data. The research furthermore suggests that activation of the visual brain areas just after the stimuli is lower in sticky mind wandering than in non-sticky mind wandering.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Vugt, M.K. van and Jin, Y.
Degree programme:	Artificial Intelligence
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	13 Oct 2022 12:49
Last Modified:	13 Oct 2022 12:49
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/23977

Actions (login required)

View Item