Application Research of Educational Data Prediction Model Based on LS-SVM in Educational Evaluation and Decision-Making

Under the promotion of the present big data technology, the analysis and research of educational data is more important to the law of educational development and decision-making. The mining of deep hidden information in education can improve problems in the education process and provide support for decision-making by relevant education departments. In response to the above issues, this study constructed an education institution evaluation model built on support vector machine library and an education data prediction model based on least squares algorithm for the evaluation and decision-making of intelligent education. The results prove that the designed model has the largest error in the prediction data in 2017, with a relative error of 0.4161%, and the relative errors of the prediction data in other years are all less than 0.05%. In summary, the proposed method has better performance and is very suitable for practical educational evaluation and decision-making applications.