A Machine Learning Approach to Predicting and Analyzing ICT Adoption Challenges in Secondary Schools of North Western Tigray, Ethiopia

Abstract

Though its application in impoverished and conflict-affected areas continues to face obstacles, information and communication technology (ICT) has emerged as a revolutionary force in global education. During the post-conflict rehabilitation phase, this study examines the obstacles to ICT adoption in public secondary schools in Tigray, Ethiopia’s North West Zone. The study combines descriptive analysis and prediction modeling using a dataset of 30,000 records gathered from 12 woredas via surveys, interviews, and administrative sources. The study uses the Linear Regression, Random Forest, and Gradient Boosting algorithms to create a machine learning-based predictive analytics framework under the direction of the Design Science Research Methodology (DSRM). To guarantee robustness and interpretability, the models were assessed using metrics including Mean Squared Error (MSE), Mean Absolute Error (MAE), and R2, in addition to feature selection and dimensionality reduction strategies. The results show that the best indicators of ICT usage frequency are teacher ICT training hours, internet connectivity, and electrical dependability, while administrative support and infrastructure deficiencies were identified as major obstacles. The best prediction performance (R2 ≈ 0.43) was obtained by Random Forest models on original characteristics, indicating their usefulness in capturing intricate, nonlinear relationships in the educational setting. By connecting educational theory with computer modeling, the study makes a theoretical and practical contribution by providing evidence-based insights to help stakeholders, school administrators, and policymakers create focused interventions. In order to guarantee sustained ICT integration in Ethiopia’s secondary education sector, this research ultimately emphasizes the necessity of integrated strategies that connect infrastructure, human ability, and leadership.