An efficient medical data classification using oppositional fruit fly optimization and modified kernel ridge regression algorithm

Medical researches utilize data mining techniques for several years and have been well known to be successful one. In the medical data have certain characteristic that make their analysis very challenging and attractive. In the proposed medical data classification research, it contains relevant and irrelevant features. Here the irrelevant features to be reduced with the aid of oppositional fruit fly optimization algorithm. In the feature selection phase the optimal subset of features are finally divided into training and testing files. The output of training and testing files is given into classifier. This classification is to be performed with the aid of Modified Kernel Ridge Regression (MKRR). KRR gets knowledge about a linear function in the space induced by the respective kernel and the data. For MKRR non-linear kernels, this corresponds to a non-linear function in the original space. The form of the model acquire knowledge by Kernel Ridge is alike to support vector regression.