Impact of Physiological Sensor Variance on Machine Learning Algorithms
Bibliographic record
Abstract
Machine learning based acute stress detection systems use physiological sensor data to objectively predict acute stress. However, machine learning algorithms developed for stress detection do not consider how machine learning algorithm performance may be affected based on a change(s) in the deployment environment. In this study, the deployment environment changes that are investigated are sensor type and sensor placement. Electrodermal activity (EDA) and skin temperature (TEMP) data from two different sensors, the RespiBAN Professional (RespiBAN) and the Empatica E4 are used to train three different machine learning models. The RespiBAN records the EDA data from the rectus abdominis and records the skin TEMP data from the sternum. The Empatica E4 sensor records both EDA and skin TEMP data from the wrist. Three different support vector machine (SVM) models were trained to classify no-stress versus stress states using EDA and skin TEMP data. The first model was trained using data from the RespiBAN wearable sensor (SVM-R), the second model was trained using data from the Empatica E4 sensor (SVM-E) and third model was trained using data from both sensors (SVM-RE). The accuracy of SVM-R on a test set recorded by the RespiBAN sensor was 100%. The accuracy of SVM-E on a test set recorded by the Empatica E4 sensor was 99%. The accuracy of SVM-RE on a test set recorded by both the RespiBAN and Empatica E4 sensor was 82%. The accuracy of the SVM-R on a test set recorded by the Empatica E4 was 64%. These results suggest that research and development cannot be hardware or placement agnostic with wearable sensing data. Sensor type and placement must be taken into consideration when reporting performance metrics of physiological based stress detection machine learning algorithms.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".