ISSN: 2182-2069 (printed) / ISSN: 2182-2077 (online)
OptCatB: Optuna Hyperparameter Optimization Model to Forecast the Educational Proficiency of Immigrant Students based on CatBoost Regression
A person's poor educational performance or academic success would hinder the struggle against poverty that plagues humanity, particularly for children who are close to completing high school. This study examines the PISA dataset containing immigrant student information from nations like the UAE, New Zealand, Canada, Qatar, Spain, and Australia. As a result, they place a high priority on analyzing the performance of immigrant students to provide them with a high-quality education. Based on the data analysis and interpretation of the findings, factors like early arrival, late arrival, wealth factors, family circumstances, and a multitude of other socioeconomic factors have an influence on the performance of students in reading, math, and science scores. The proposed OptCatB model makes predictions regarding the academic success of immigrant students by applying an optimized CatBoost regressor by keeping reading, math, and science as target variables. We trained the model using the optimized parameters after tuning the hyperparameters of the CatBoost algorithm by using a hyperparameter optimization technique termed Optuna. The OptCatB model outperformed compared to the other selected regression models with RMSE of 54.231, MAE of 43.104, MAPE of 9.931 and RSE of 0.54