Teenager Depression Diagnosis Using Physical and Algorithmic Deep Learning Methods (RNN)

Authors

  • Minjae Kim Choate Rosemary Hall

DOI:

https://doi.org/10.26821/IJSHRE.10.8.2022.100803

Abstract

Recently, the suicide rate among teenagers has risen to the point that it has become a massive societal issue. Suicide stems from depression; as such, it is critical to diagnose depression in children and adolescents early and to address the mental issue promptly. However, it is difficult for people to realize that what they’re going through is indeed depression, leading to many people developing depression without knowing what is happening. In this research paper, I propose a method of detecting early signs of depression that takes both physical and mental factors into account. Depression causes physical changes within the body in a variety of ways. The paper discusses which physical devices can be introduced to measure these changes. There are also various forms and tests that may be used to “grade” a person on a standardized scale of depression. Since patients must manually respond to the questions, I have categorized this procedure as a physical system. We may deduce the severity of the condition from the information that individuals submit naturally, in addition to obtaining responses from them. Deep learning can be employed to carry out this procedure. To be more precise, I applied RNN to implement binary text classification. The evaluation used real data to demonstrate its practicality. The study closes with suggestions for enhancing the system.

References

. K Kroneke et al, The PHQ‐9: validity of a brief depression severity measure, 2001

. AT Beck et al, Beck depression inventory (BDI-II), 1996

. JT Biggs et al, Validity of the Zung self-rating depression scale, 1978

. T Mikolov, Recurrent neural network based language model, 2010

. P Liu, Recurrent neural network for text classification with multi-task learning, 2016

Downloads

Published

2022-08-13

How to Cite

Minjae Kim. (2022). Teenager Depression Diagnosis Using Physical and Algorithmic Deep Learning Methods (RNN) . iJournals:International Journal of Software & Hardware Research in Engineering ISSN:2347-4890, 10(8). https://doi.org/10.26821/IJSHRE.10.8.2022.100803