This thesis is the application of data mining techniques to develop a system to predict the choice of higher education. Using a decision tree algorithm C4.5 (or J48) and ID3 algorithm modeling. By sharing information with how to monitor cross (k-Fold Cross Validation) A 5-Fold Cross Validation and the 10-Fold Cross Validation, and how to divide the random data by dividing the percentage (Percentage Split) by sharing information kit learn. and test in 80:20, 70:30 and 34:66 making model predictions of 10 models and compare the accuracy of each model. The results showed that the algorithm J48 way to share information and monitor cross-type 5-Fold and 10-Fold with accuracy in the classification of 88.9% and 81.35%, and the share random information with the division. 20,30, and 66 percent have an accuracy of 97.43%, 82.22% are classified and 91.71% respectively algorithm ID3 way to share information and monitor cross-type 5-Fold 10-Fold in value. the accuracy of the model is 95.73% and 83.05% and share data with random breaks 20,30 and 66 percent have an accuracy of 93.16%, 83.89%, and the model is 95.34%, respectively.
การแปล กรุณารอสักครู่..
