Evaluation and Prediction Algorithm of Students' Comprehensive Quality Based on Big Data and Machine Learning

The purpose of this study is to explore the evaluation and prediction algorithm of students' comprehensive quality based on big data and machine learning technology. This paper designs a multi-task neural network (MNN) model, which combines several evaluation indexes, such as academic performance, social activities, health status, etc., to comprehensively evaluate students' comprehensive quality level. Through the experimental verification of real student data, it is found that MNN model has better performance than the traditional benchmark model in predicting students' comprehensive quality, showing lower Mean absolute error (MAE) and mean-squared error (MSE). In addition, this study also analyzes the feature importance of the model, and finds that study time and social activities are the two factors that have the greatest influence on the comprehensive quality evaluation. The experimental results also show that our model has good performance stability and generalization ability on different data subsets. This study provides new ideas and methods for students' evaluation in the field of education, and provides an important reference for schools and education management departments to formulate personalized education policies.