A Non-invasive Mental Health Risk Predictor Using Machine Learning Models Utilising Music Listening Habits

Authors

Abstract

Globally, there has been a rise in mental health issues such as insomnia, anxiety, and depression. However, the stigma that is associated with such a diagnosis makes individuals not want to seek help. Recent research has explored the relationship between music listening habits and mental health status, offering promising insights into the potential of leveraging this data for predictive modelling. This research proposes a non-invasive approach that integrates features extracted from music listening patterns including demographic and lifestyle data to build machine learning models that detect mental health conditions such as insomnia, depression and anxiety levels The results show that Random Forest achieved an accuracy of 76.35%, which highlights the potential of using music listening habits to predict mental health states. The findings of this study provide valuable insight into the relationship between music and mental health predictors- namely depression, anxiety and insomnia across different age groups.

Recommended Citation

(2024) "A Non-invasive Mental Health Risk Predictor Using Machine Learning Models Utilising Music Listening Habits," Communications of the IIMA: Vol. 22: Iss. 1, Article 8.
DOI: https://doi.org/10.58729/1941-6687.1476
Available at: https://scholarworks.lib.csusb.edu/ciima/vol22/iss1/8