This Dissertation Studies the Results of the Postgraduate Entrance Examinations Based on Clustering and Random Forest Hybrid Algorithm
As the number of graduate school entrance examination candidates increases annually and the actual admission rate continues to decline,this study aims to assist students in planning their path for these exams rationally.Initially,the study employs machine learning techniques to select,preprocess,and normalize original academic performance data.Subsequently,the K-means clustering algorithm is ap-plied to reduce the volume of data input into the predictive model and to enhance the quality of the samples.By utilizing Principal Component Analysis(PCA)for dimensionality reduction,the study effec-tively minimizes interference among variables in the dataset,thereby enhancing computational efficiency.Additionally,PCA aids in reducing overfitting in the predictive model during training,achieving a sub-stantial reduction in the number of features in the dataset.Ultimately,the Random Forest(RF)algo-rithm is utilized to generate predictive results.The findings indicate that the accuracy of this predictive model exceeds 86.5%,providing a highly valuable reference for the objective prediction of success or failure in the graduate school entrance examinations.
万方数据
维普
