Open AccessPrediction of Stress and Mood using Neural Network, LSTM and Transfer learning
Author(s) -
V T Neethu
Publication year - 2021
Publication title -
international journal for research in applied science and engineering technology
Language(s) - English
Resource type - Journals
ISSN - 2321-9653
DOI - 10.22214/ijraset.2021.35347
Subject(s) - mood , task (project management) , feeling , transfer of learning , stress (linguistics) , affect (linguistics) , computer science , artificial neural network , artificial intelligence , cognitive psychology , psychology , social psychology , communication , linguistics , philosophy , management , economics
Stress can be a feeling of emotional or physical tension. It can come from any thought or event that makes you feel frustrated, disturbed, angry or nervous. It also affect the mood of the person. This study is conducting to predict the stress and mood based on heart rate variability which can be collected using Fitbit devices or Apple watches nowadays. In this work SWELL dataset available from the Kaggle repository is used. Neural Network and LSTM is used to predict the stress and mood. Predicting the stress is considered as first task and as mood prediction as second task. For second task prediction, the model created for first task is reused as pretrained model where we make use of transfer learning.