Volume 18, No. 5, 2021
SEDDA-An Optimal Feature Selection Approach For Breast Cancer Diagnosis
S. Vani Kumari , K. Usha Rani
Abstract
Soft Computing is an area of Computer Science that aims at solving different types of problems ranging from Medical Image Classification, Content-Based Image Retrieval, Sentiment Analysis, etc. Medical Image Classification involves extracting the essential information from the images and develop models to classify the images. With the advent of the latest technologies in the medical field, there is an explosion of high dimensional data that is to be processed. Hence techniques are required to find the important features of data that are to be processed which can be achieved by feature selection. Feature selection is being applied in fields like Data Mining and Data Science. Feature Selection aims at the elimination of irrelevant and redundant features that adversely affect the performance of any machine learning technique. This paper aims to propose a novel metaheuristic method named Separated Enemy Driven Dragonfly Algorithm (SEDDA) which is an improvisation of the Dragonfly algorithm for selecting an optimal subset of features that are extracted from digital mammograms. The texture features of the segmented area of interest have been extracted and the optimal feature set has been obtained using SEDDA. An accuracy of 94.5 and sensitivity 87.3was achieved by Multi-Layer Perceptron with Back Propagation which considered for classification purpose. The performance of the novel algorithm is compared with genetic algorithm, particle swarm optimization, ant colony with cuckoo search and dragonfly algorithm. The results show that the novel algorithm is more accurate than the other algorithms.
Pages: 226-242
Keywords: Soft Computing, Metaheuristic , Dragonfly algorithm,