Classification of Breast Cancer using Fast Fuzzy Clustering based on Kernel

Breast cancer is the second leading cause of death in women in the world. The classification is the initial process of executing patient treatment, which is important as it increases life expectancy as well as quality. In this paper, a new method is proposed based on kernel, which is modified from KC-Means: it combines K-Means, Fuzzy C-Means algorithm, and kernel function. The C-Means algorithm is applied on the centers of a fixed number of groups founded by K-Means, and the kernel function is expected to improve the accuracy of classification with its ability to separate data which cannot be separated linearly. We applied the proposed method on a dataset of 201 breast cancer and 85 non-breast cancer samples from the UC Irvine Machine Learning Repository. Results concluded that fast fuzzy clustering has an accuracy of 85.26%, but fast fuzzy clustering based on kernel is 89.74%, with a better running time on average than 90.95% with the same method.