Evaluation of Stress Based on Multiple Distinct Modalities Using Machine Learning Techniques - Proceedings Study

Aug 5, 2026

Upload by Violet

The study addresses the complexity and time cost of predicting working professionals’ stress and the limitations of earlier machine-learning approaches, including design complexity, frequent misclassification, and reduced efficiency. A deep learning pipeline is proposed that integrates convolutional neural networks for stress classification with dataset preprocessing, feature extraction, and optimal feature selection via principal component analysis (PCA). Preprocessing removes duplicate characteristics and fills missing values to improve signal quality and model reliability.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Introduction

Stress impact and detection need

AI models and application contexts

Physiological mechanisms and signals

Methods and proposed approach

CNN-based classification with preprocessing

Feature extraction and PCA-based feature selection

Evaluation of stress based on multiple distinct modalities using machine learning techniques Samarendra Narayana Pradhan1, Reddy Shiva Shankar1,2, Shekharesh Barik1, Bhabodeepika Mohanty1, Venkata Rama Maheswara Rao1,3 1Department of Computer Science and Engineering, Faculty ofCSE, Dhaneswar Rath Institute of Engineering and Management Studies (DRIEMS), Tangi, Odisha, India 2Faculty ofCSE, Sagi Rama Krishnam Raju Engineering College (A), Bhimavaram, Andhra Pradesh, India 3Faculty ofCSE, Shri Vishnu Engineering College for Women (A), Bhimavaram, Andhra Pradesh, India Article history: Received Jun 22, 2023 Revised Oct 23, 2023 Accepted Nov 2, 2023 Keywords: Artificial intelligence Auto encoders Convolutional neural networks Deep learning Machine learning Nowadays, one of the most time-consuming and complex study subjects is predicting working professionals' stress levels. It is thus crucial to estimate active professionals' stress levels to aid their professional development. Several machine learning (ML) and deep learning (DL) methods have been created in earlier articles for this goal. But they also have drawbacks, such as increased design complexity, a high rate of misclassification, a high incidence of mistakes, and reduced efficiency. Considering these issues, the objective of this study is to make a prognosis about the stress levels experienced by working professionals by using a cutting-edge deep learning model known as the convolutional neural networks (CNN) . In this paper, we offer a model that combines CNN-based classification with dataset preprocessing, feature extraction, and optimum feature selection using principal component analysis (PCA) . When the raw data is preprocessed, duplicate characteristics are eliminated, and missing values are filled. Principal component analysis This is an open access article under the CC BY-SA license. Corresponding Author: Samarendra Narayana Pradhan Department of Computer Science and Engineering, Dhaneswar Rath Institute of Engineering and Management Studies (DRIEMS) Tangi, Cuttack, Odisha, India Email: [sunilcst263@gmail.com](sunilcst263@gmail.com) Article Info ABSTRACT 1. INTRODUCTION Stress affects more than 70% of the population. Long-term stress causes reduced immunity, cancer, cardiovascular disease, depression, diabetes, and drug addiction [1] . Stress damages mental and physical health. Developing trustworthy technologies to identify human stress quickly is crucial. These technologies might detect stress continuously. Hence, stress may be reduced by regulating daily activities, and healthcare practitioners can better manage stress-related disorders. Researchers have developed many stress-detecting methods. The sympathetic nervous system releases adrenaline and cortisol when threatened [2] . So, this condition may significantly damage a stressed person's daily life and health [3] . Hence, stress and other factors may cause fatal car accidents [4] . According to the World Health Organization (WHO), 1.35 million road traffic crashes kill 5–29-year-olds [5] . So, it is important to create strategies to immediately spot stress in drivers so they may avoid automobile accidents and injuries, which can greatly impact the driver's and accident victims' lives. These artificial intelligence (AI) models may assess stress in a variety of contexts, such as the workplace [6], while driving [7], in the classroom [8], and during crises [9] . AI models may help self-regulate stress, and extreme stress might alert human resources (HR), management, or teachers to change the atmosphere and workload. AI models may help with affective computing in other soft skills. Stress may cause depression, addiction, and cardiovascular diseases. Emotional tension now harms mental and physical health. Psychological stress is mitigated by ecological momentary assessment (EMA) [10] . Nevertheless, the two systems require real-time psychological stress monitorin

FAQs

Why is stress prediction for working professionals considered challenging?

It is described as time-consuming and complex, and earlier ML/DL approaches may suffer from design complexity, misclassification, and reduced efficiency.

What is the main method proposed in the study?

The approach combines CNN-based classification with dataset preprocessing, feature extraction, and optimal feature selection using principal component analysis (PCA).

How does the study prepare the dataset before modeling?

Duplicate characteristics are eliminated during preprocessing, and missing values are filled to improve data quality before training the model.

Aug 5, 2026

Upload by Violet