Design of the Prediction Model for Adolescents’ Stress Using Deep Learning
- Authors
- Lee, H.; Jung, E.
- Issue Date
- May-2021
- Publisher
- Springer Science and Business Media Deutschland GmbH
- Keywords
- Adolescents’ stress; Deep learning; kNN; Prediction
- Citation
- Lecture Notes in Electrical Engineering, v.782, pp 23 - 29
- Pages
- 7
- Journal Title
- Lecture Notes in Electrical Engineering
- Volume
- 782
- Start Page
- 23
- End Page
- 29
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/62431
- DOI
- 10.1007/978-981-16-4118-3_3
- ISSN
- 1876-1100
1876-1119
- Abstract
- Adolescents are exposed to various types of stress from parents, study, or friends in their school life. Though these stresses cannot be avoided, the proper monitoring of affecting variables can give educators a chance to help their youths to ease the stresses. Deep Learning is superior to other machine learning methods in terms of prediction performance, but it has a weakness to explain the effects of variables due to its black box characteristic. In addition to this, it requires all input variables to use the trained neural network, but it is frequently impractical to gather all variables in the actual education field. To resolve this issue, the authors suggest the design approach combining Deep Learning and kNN. The authors use feature importance with kNN and reduce variables into one third, but the result of performance evaluation shows that the approach can save the advantage of prediction performance of Deep Learning while reducing the number of variables. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - ETC > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/62431)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.