Abstract
the most doctors spend a lot of time for detecting a benign tissue which is easily be distinguished from malignant one in a computerized community. This denotes to a waste of time and resources that can be spent in classifying the difficult cases. As a result, many researchers began to develop diagnostic methods with aid of computer applications that uses image processing techniques. It helps to classify existence of diseases such as breast cancer where a lot of features are used to distinguish this disease. This paper employees a meta-heuristic algorithm (Dolphin Echolocation Algorithm DEA) to select the most effective features from all expensive used features to accurate and fast classification of breast cancer. In this research, Fine Needle Aspiration images are used. Additionally, three classifiers (SVM, BP-NN, and KNN) are utilized to classify medical data. For increasing the accuracy and reducing the time, many feature selection algorithms are used. The results show that meta-heuristic algorithms (GA and suggested DEA) are outperformed other feature selection methods.
Keywords
Breast Cancer; Feature Selection; Feature extraction; Classification; FNA.