Prediction of Mind-Wandering with Electroencephalogram and Non-linear Regression Modeling
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 limitations of existing methods for measuring mind-wandering (MW), which typically rely on self-reporting or binary classification models that fail to capture the continuous intensity of the phenomenon. While MW is linked to various psychological issues, including depression and anxiety, current physiological prediction models using fMRI or eye-tracking are often impractical for general use due to cost or situational constraints. The authors propose that electroencephalogram (EEG) data, combined with non-linear regression modeling, can effectively predict MW intensity. This approach aims to provide a high-temporal-resolution metric that is applicable in diverse settings, including closed-eye states like meditation, and facilitates the development of portable neuro-feedback devices. The researchers recorded EEG data from 50 participants performing a Sustained Attention to Response Task (SART). MW intensity was measured using probe-caught thought sampling, where participants rated their focus on a 7-point Likert scale every 20 seconds. EEG signals were processed to extract power and coherence values across eight frequency bands from 17 electrodes. The dataset was split into training and test sets, with predictors selected based on Pearson’s correlation coefficients. The study employed Support Vector Machine Regression (SVR) to fit five models: two non-linear models (using Radial Basis Function kernels) and two linear models, with variations in the number of electrodes used to assess model versatility. One model used all 16 selected electrodes, while another used only eight electrodes to simulate limited-measurement environments. Results indicated that all SVR models significantly outperformed a single-variable linear regression model. The best-performing non-linear model using 16 electrodes achieved a correlation coefficient of r = 0.54 between estimated and measured MW intensity. The non-linear model using only eight electrodes achieved r = 0.49, which was significantly higher than the corresponding linear model (r = 0.39). These findings confirm that non-linear SVR models provide superior predictive precision compared to linear alternatives, particularly when using a limited number of electrodes. The study validates the use of EEG coherence and power features as robust indicators of MW intensity. The significance of this work lies in demonstrating that MW intensity can be reliably predicted using EEG data and non-linear machine learning algorithms. By showing that a small subset of electrodes can yield accurate predictions, the study supports the feasibility of developing simplified, portable EEG devices for real-time neuro-feedback. This advancement allows for the investigation of MW dynamics in previously inaccessible contexts, such as sleep or meditation, and offers a tool for monitoring time-series variations in attention, potentially aiding in the treatment of psychiatric conditions associated with excessive mind-wandering.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | canonical_url | — | — | 1 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| enrich | success | semantic_scholar | — | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 10 | 2026-08-11 |
| verify | success | — | — | — | 2 | 2026-08-10 |
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.