Author/Authors :
gümüşçü, abdülkadir harran üniversitesi - mühendislik fakültesi - elektrik-elektronik mühendisliği bölümü, Turkey , taşaltın, ramazan harran üniversitesi - mühendislik fakültesi - elektrik-elektronik mühendisliği bölümü, Turkey , aydilek, ibrahim berkan harran üniversitesi - mühendislik fakültesi - bilgisayar mühendisliği bölümü, Turkey
Title Of Article :
C4.5 Decision tree pruning using genetic algorithm
Abstract :
Decision tree is a machine learning algorithm that is used for classification and regression. Many approaches were proposed to build decision trees. C4.5 decision tree that is one of these approaches, is frequently used in many fields. Large number of attributes of the data set that is used for building decision tree causes unnecessary branches and nodes on decision tree. Unnecessary branches and nodes cause overfitting. Overfitting negatively affects classification success rate. In this paper, a novel pruning algorithm is proposed to reduce the effects of overfitting. Successful results were obtained by optimizing confidence factor (CF) of C4.5 algorithm executed in Weka using genetic algorithm.
NaturalLanguageKeyword :
Genetic algorithm , Decision tree , Pruning
JournalTitle :
dicle university journal of institute of natural and applied science
