Comparing ID3, C4.5 and CART Decision Tree algorithms

Description

Decision Trees are one of the popular data mining classifiers. The Decision Tree algorithms are supervised learning algorithms. These algorithms work on training samples and prepare the trees. using the tree new dataset instances are classified to get their class labels. In this presented work, a simulation has been made for comparing the performance of three decision tree algorithms namely ID3, C4.5, and CART algorithms. Thus, the dataset in comma-separated values (CSV) format has collected from the UCI repository [1]. to classify them WEKA [2], is implemented for providing GUI. Users can select a dataset and classify them into their classes. during experiments of algorithms, the performance of the algorithm on different datasets has been measured and demonstrated in a line graph. To demonstrate the comparative performance of decision trees the precision, recall, f1-score, time consumption, and memory usage are measured.

References:

[1]https://archive.ics.uci.edu/ml/index.php

[2]https://www.cs.waikato.ac.nz/ml/weka/