Ateş Böceği Algoritması Destekli Aşırı Öğrenme Makinesi ile Göğüs Kanseri Veri Kümelerinin Sınıflandırılması

Abstract

The Breast cancer is the second cancer type which causes death of women. The premature detection of cancer and the suitable treatment applied to cancer cells can reduce the deadly risk. The medical doctors can make faults in diagnosis of the cancer disease. The performance of artificial intelligence methods (AIMs) containing increased thanks to rapid improvements in the technologies of the computer hardware. AIMs can be used regarding the enhancement of diagnostic accuracy. Standard Gradient–Based back propagation artificial neural networks (BP–ANN) has been commonly utilized in the diagnosis of the breast cancer disease. Even though BP–ANN are good performance in diagnosis of cancer disease, it has some limitations such as possible to be trapped in local minima and long time in the training process. In this study, the extreme learning machine assisted by heuristic firefly algorithm (FF–ELM) is proposed for diagnoses of breast cancer disease on the Breast Cancer Wisconsin Dataset. The diagnostic performance of proposed FF–ELM was compared with the standard ELM and BP–ANN methods. The results show that FF–ELM provides a meaningful enhancement regarding the classification performance and it can be used as a powerful technique for the medical problems.