Open AccessUtilizing the Genetic Algorithm to Pruning the C4.5 Decision Tree Algorithm
Author(s) -
Maad M. Mijwil,
Rana Ali Abttan
Publication year - 2021
Publication title -
asian journal of applied sciences
Language(s) - English
Resource type - Journals
ISSN - 2321-0893
DOI - 10.24203/ajas.v9i1.6503
Subject(s) - overfitting , machine learning , decision tree , computer science , artificial intelligence , classifier (uml) , pruning , decision tree learning , incremental decision tree , id3 algorithm , genetic algorithm , algorithm , data mining , artificial neural network , agronomy , biology
A decision tree (DTs) is one of the most popular machine learning algorithms that divide data repeatedly to form groups or classes. It is a supervised learning algorithm that can be used on discrete or continuous data for classification or regression. The most traditional classifier in this algorithm is the C4.5 decision tree, which is the point of this research. This classifier has the advantage of building a vast data set and does not stop until it reaches the desired goal. The problem with this classifier is that there are unnecessary nodes and branches leading to overfitting. This overfitting can negatively affect the classification process. In this context, the authors suggest utilizing a genetic algorithm to prune the effect of overfitting. This dataset study consists of four datasets: IRIS, Car Evaluation, GLASS, and WINE collected from UC Irvine (UCI) machine learning repository. The experimental results have confirmed the effectiveness of the genetic algorithm in pruning the effect of overfitting on the four datasets and optimizing confidence factor (CF) of the C4.5 decision tree. The proposed method has reached about 92% accuracy in this work.