The Developed of Models to Forecast Aptitude of Learning by 4 MAT theory Using the Decision Tree Analysis

Abstract

The objective of this research was to forecast aptitude of learning by 4 MAT theory using the analyzing with a decision tree. To build and test the model predictions a researcher using WEKA on modeling techniques with a decision tree was used. The investigator used the method of classification and learning methods by J48 (algorithm C4.5 Version 8.0). To learn from the Train Dataset and the modeling of trees to be used for the classification of information as the algorithm web continues to evolve as a module on the network. The model will be available in the form of rules, the classification of learning with the training set then put to the test data utilizing the k-fold cross-validation and percentage split. The results showed that forecast aptitude of learning by 4 MAT theory using the analyzing with a decision tree that was developed the training set and share data sets from each a higher performance model prepared in other ways. The accuracy of 77.65 % and 77.90 % accuracy equal to 77.65 % of the memorial and the balance was 77.40 %. Showing how training set and share data sets can be used to develop the forecast aptitude of learning by 4 MAT theory using the analyzing with decision tree with accuracy and precision in the prediction ability of learning by 4 MAT method as well.