Analysis Of The K12 Education Of United States Using Machine Learning And Data Mining Techniques

Quality K-12 education is essential for a student’s success in college and future career. In this paper, we proposed a Random Forest-based algorithm to identify the principal factors affecting the academic success of students in K-12 education across various states in the United States. Our model outperforms other machine learning-based models like Lasso and ElasticNet, reaching an MSE of 0.910. Grid search is used to automatically search for a set of optimal hyperparameters and we concluded that enrollment number, the year when data is recorded and capital outlay expenditure per student have the strongest effect on student’s academic success.