Javascript must be enabled for the correct page display

Predicting well-being with wearable sensor data

Drijfhout, Desmond (2019) Predicting well-being with wearable sensor data. Bachelor's Thesis, Artificial Intelligence.


Download (764kB) | Preview
[img] Text
Restricted to Registered users only

Download (124kB)


The global population is ageing rapidly. The medical industry faces challenges such as overburdened health-care delivery systems and growth in physician demand. One solution that could potentially help handle the ageing population more efficiently is the automatic prediction of well-being on the basis of different well-being predictors. This could allow for more targeted care for patients. To examine whether it is possible to develop automatic prediction of well-being, we gathered data on sleep, activity, and time away from home for 22 elderly participants living in The Netherlands for a period of 10 days, using a medical sensor watch. In addition, we gathered self-reported assessments of well-being from the participants as well as nurse evaluations on the participants’ well-being. We used and compared different types of classifiers to try to automatically predict well-being from the sensor data. We tried feature selection and different re-sampling techniques to improve the models. We show that, using the Random Forest classifier, both the self-reported assessments (after feature selection) and nurse evaluations (after re-sampling) can be predicted with respectively 94% accuracy (92.75% sensitivity) and 90.64% accuracy (90.14% sensitivity). Furthermore, we show that manually assessing the well-being of a patient solely based on looking at the sensor data is not a good predictor of the self-reported assessments nor the nurse evaluations. Last, we show potential problems, like overfitting, that we encountered collecting and evaluating the results.

Item Type: Thesis (Bachelor's Thesis)
Supervisor nameSupervisor E mail
Vugt, M.K.
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 13 Feb 2019
Last Modified: 15 Feb 2019 15:36

Actions (login required)

View Item View Item