ItemArticle

Early detection of mental health disorders using machine learning models: An analysis based on behavioral and voice data

Title
Early detection of mental health disorders using machine learning models: An analysis based on behavioral and voice data
Creator
Vats, Prashant; Tak, Tan Kuan; Upreti, Kamal; Mahajan, Shubham; Kshirsagar, Pravin R.; Upadhyay, Govind Murari
Description
Mental illnesses are to be detected promptly and correctly to intervene effectively and in time. In this paper, a multi-stage NeuroVibeNet model of early mental disorders detection based on multimodal behavioral and voice data is proposed. It starts with the preprocessing of data that is high-quality and consistent, such as mean imputation, min-max normalization, outlier detection, noise reduction, and short-time energy extraction. The majority of the advanced methods employed in extracting temporal, spectral, and complex features include multiscale entropy, soft dynamic time warping, spectral contrast analysis, formant frequency analysis, and a one-dimensional convolutional neural network autoencoder. The feature selection is done via a sparse autoencoder that is used to maximize relevance and minimize redundancy. The chosen features are fed into the NeuroVibeNet architecture, where TabNet is used to process behavioral data, and Capsule Networks are used to process voice data to allow learning representations with attention and hierarchy. Lastly, a voting-based ensemble classifier uses the two modalities to combine the predictions to make strong classification decisions. The structure is coded in Python and tested on three benchmark datasets with the accuracy of 0.9839, 0.9856, and 0.9855, which is better than the current approaches. Copyright 2026. Published by Elsevier Ltd.
Source
Computers and Electrical Engineering;Volume;132;Issue;;Article No.;110996;
Date
01-01-2026
Publisher
Elsevier Ltd
Subject
Behavioral data; Capsule network; Feature selection; Median imputation; Noise reduction; Tabular neural network; Voice data
Coverage
Vats P., Department of Computer Science and Engineering, Manipal University Jaipur, Rajasthan, Jaipur, 303007, India; Tak T.K., Engineering Cluster, Singapore Institute of Technology, 828608, Singapore; Upreti K., Department of Computer Science, CHRIST (Deemed to be University), Uttar Pradesh, Delhi NCR, 201003, India; Mahajan S., Amity School of Engineering and Technology (ASET), Amity University, Haryana, Gurugram, 122413, India; Kshirsagar P.R., Electronics & Telecommunication Engineering, J D College of Engineering & Management, Maharashtra, Nagpur, 441501, India; Upadhyay G.M., Department of Computer Applications, Manipal University Jaipur, Rajasthan, Jaipur, 303007, India
Rights
Restricted Access; Hardcopy may be available in the library
Relation
ISSN: 457906; CODEN: CPEEB
Format
online
Language
English
Type
Article

Citation

Vats, Prashant; Tak, Tan Kuan; Upreti, Kamal; Mahajan, Shubham; Kshirsagar, Pravin R.; Upadhyay, Govind Murari, “Early detection of mental health disorders using machine learning models: An analysis based on behavioral and voice data,” CHRIST (Deemed To Be University) Institutional Repository, accessed June 18, 2026, https://archives.christuniversity.in/items/show/22234.

Output Formats