Hybrid ABC and black hole algorithm with genetic operators optimized SVM ensemble based diagnosis of breast cancer.

Forever and a day, breast cancer has caused significant negative impacts on the quality of lives of number of women, more often than not turning into a fatal disease. The growth in the number of such cases has constantly been a major concern for the community as well as medical experts. To prevent irreversible damages caused by the disease, early identification of breast cancer is essential. Various researches and techniques have been devised in the past as an attempt to achieve this task with appreciable accuracy. As an advancement to these pre-existing algorithms and methods, we have devised a model by exploiting the techniques of nature-inspired metaheuristics in order to efficiently detect breast cancer at an early stage while maintaining acceptable levels of accuracy. In this paper, we propose a hybrid model, namely "hybrid artificial bee colony and black hole with genetic operators (GBHABC)", for the early detection of breast cancer. In the proposed model, we employed a support vector machine (SVM) ensemble technique, optimized using the proposed GBHABC model. This model combines the techniques of two major algorithms, namely artificial bee colony (ABC) and black hole (BH), guided through crossover and mutation genetic operators. Datasets from the well-known UCI breast cancer repository have been used to train the models and evaluate test result. For a fair and accurate evaluation of the model, a number of metrics have been examined including accuracy, sensitivity, specificity, F1-score and precision. An impeccable accuracy of 99.42% was obtained on the UCI dataset, clearly outperforming any literature in the same field. [ABSTRACT FROM AUTHOR]