Optimized Predictive Models in Health Care Using Machine Learning

3Improving Accuracy in Predicting Stress Levels of Working Women Using Convolutional Neural Networks

Purude Vaishali Narayanro^1,2, Regula Srilakshmi¹, M. Deepika¹ and P. Lalitha Surya Kumari²*

¹Department of CSE, Neil Gogte Institute of Technology, Hyderabad, Telangana, India

²Department of CSE, Koneru Lakshmaiah Education Foundation, Hyderabad, Telangana, India

Abstract

Currently, the world is facing a severe and prevalent issue called stress, which significantly impacts women’s health and the development of their children. To aid working women in their professional and personal growth, assessing their stress levels accurately is crucial. Artificial intelligence (AI) algorithms have been used for stress level prediction. However, these models are prone to misclassification and errors, and their design can be complicated and less efficient. To overcome the limitations, we propose a convolutional neural network (CNN) model to identify stress levels in working women. Our framework includes creating a dataset, extracting the features, selecting optimal features, and binary classification using CNNs. We also handle missing values and remove duplicate attributes during data preprocessing. Our approach using CNNs is expected to outperform previous ML and DL algorithms in accurately predicting stress levels in working women.

Keywords: Stress prediction, working women, classification, CNN

3.1 Introduction

Predictive modeling uses statistics and machine learning algorithms to create ...