Performance Assessment of Certain Machine Learning Models for Predicting the Major Depressive Disorder among IT...

Area of Science:

• Neuroscience
• Computer Science
• Psychiatry

Background:

• Major Depressive Disorder (MDD) is a prevalent mental health condition affecting individuals globally.
• Many individuals with MDD do not seek professional help, and early identification remains a challenge.
• Depression often co-occurs with anxiety, particularly in high-pressure work environments like the IT sector.

Purpose of the Study:

• To analyze IT employees for the presence of major depressive disorder.
• To develop and evaluate an artificial neural network model for effective MDD classification.
• To compare the performance of a deep multilayered perceptron with backpropagation against other machine learning models.

Main Methods:

• Collected data from IT professionals through manual input and sensor readings.
• Implemented a multilayered neural perceptron (MLP) model.
• Utilized the backpropagation technique for model training and classification.

Main Results:

• The deep MLP model with backpropagation achieved superior performance in classifying depressed individuals.
• The developed model effectively distinguished between depressed and non-depressed individuals within the study cohort.
• The findings indicate the potential of advanced neural networks in mental health diagnostics.

Conclusions:

• Deep learning models, specifically MLP with backpropagation, offer a promising approach for accurate MDD detection.
• This study highlights the utility of technology in identifying mental health challenges in specific professional groups.
• Further research can explore integrating diverse data sources for enhanced diagnostic capabilities.