A new hybrid approach for mining breast cancer pattern using discrete particle swarm optimization and statistical method

Breast cancer is one of the leading causes of death among the women in many parts of the world. In 2007, approximately 178,480 women in the United States have been found to have invasive breast cancer. In this paper, we have developed an efficient hybrid data mining approach to separate from a population of patients who have and who do not have breast cancer. The proposed data mining approach has consists of two phases. In first phase, the statistical method will be used to pre-process the data which can eliminate the insignificant features. It can reduce the computational complexity and speed up the data mining process. In second phase, we proposed a new data mining methodology which based on the fundamental concept of the standard particle swarm optimization (PSO) namely discrete PSO. This phase aimed at creating a novel PSO in which each particle was coded in positive integer numbers and has a feasible system structure. Based on the obtained results, our proposed DPSO can improve the accuracy to 98.71%, sensitivity to 100% and specificity to 98.21%. When compared with the previous research, the proposed hybrid approach shows the improvement in both accuracy and robustness. According to the high quality of our research results, the proposed DPSO data mining algorithm can be used as the reference for making decision in hospital and provide the reference for the researchers.