Predicting Student Smoking Habits with Machine Learning Techniques

Abstract

Smoking among students remains a health concern, the intent of this work is to predict the students those who smokes cigarette by utilizing a machine learning-based approach based on behavioral, socioeconomic and other factor. The model should utilize available data to accurately classify students into smoker and non-smoker categories. The data is collected through Google forms from random people. Here in this work Random Forest models, Logistic Regression techniques, and Decision Tree methods are employed for building prediction model for smoking behavior. Comparative analysis of these three algorithms provides us the vision on which method is significant for the task