Unveiling Dropout Patterns: A Python-Powered Analysis of Student Attrition in School Education

Abstract

Student dropout rates pose a significant challenge in Gujarat, India, affecting both societal welfare and economic development. This study aims to diagnose and predict student dropout patterns using data analytics and machine learning (ML) algorithms. Extensive datasets from Gujarat's educational authorities are preprocessed with Python libraries such as Pandas and NumPy, ensuring quality and consistency. Exploratory data analysis (EDA) uncovers insights and trends, while feature engineering extracts relevant characteristics for ML models. The study implements ML algorithms Random Forest (RF), Decision Trees (DT), Logistic Regression (LR), K-Nearest Neighbors (KNN), AdaBoost, XGBoost, Support Vector Machine(SVM) to construct predictive models. The findings reveal critical factors influencing dropout rates and assess model performance using model accuracy. This research provides novel insights into student dropout rates in Gujarat, offering a comprehensive, data-driven approach to address educational attrition and guide targeted intervention strategies and policy recommendations for improving educational outcomes in the region.