Research on small sample online learning data processing and student performance prediction methods
WU Fang
WANG Qin
MAO Hongben
Abstract：Objective To improve the accuracy of student performance prediction in online teaching.Methods 161 first-year students from five classes of the Clinical Medicine School of Jiangsu Health Vocational College were selected,data augmentation,quantification of learning behavior,regression analysis,and classification analysis were used to predict the final grades of students in the basic course of imaging electronics in the spring semester.Results In the regression task,taking the support vector machine model as an example,the RMSE decreased from 13.142 before quantization to 12.388 after quantization,and decreased to 2.1123 after expansion;In binary classification tasks,accuracy,precision,recall,and F1 increased from 0.83230,0.8375,0.9926,and 0.9085 to 0.9833,0.9981,0.9722,and 0.9850,respectively;In the three classification tasks,accuracy,Macro precision,and Macro recall increased from 0.5466,0.4908,and 0.4879 to 0.9913,0.9903,and 0.9913.Conclusions The behavior quantification and data augmentation methods used in this study contribute to the construction of regression and classification models,with high accuracy in predicting final grades.
Keywords：Small sampleGaussian expansionBehavioral quantificationPerformance prediction
Publication Date：2024-11-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 502-508 )