Özet:

Feature selection contributes to the success of classifiers and the reduction of processing time by reducing data size and the selection of the best attribute set in data mining and pattern recognition applications. In this study, a feature selection algorithm has been developed by using the Artificial Bee Colony (ABC) optimization algorithm, which mimics the smart behavior of foraging honey bees. The success of the selected features has been tested using an Artificial Neural Network (NN) classifier. Taken from the UCI Machine Learning Repository, breast cancer dataset were used to demonstrate the success of the methods. Classification accuracy for this data set was obtained through using 10-fold cross-validation method. Classification accuracy of the proposed system reached 98.71% for breast cancer dataset. According to the result, proposed method performance is successful compared to other results attained.

Anahtar Kelimeler: