A Hybridized Levy Flight Fruit Fly Optimization Based Kernel Extreme Learning Machine for BioMedical Data Classification

The main motive behind classification is to map an input feature space to a predefined class labels in high dimensional microarray data sets to enhance the classification accuracy and decrease the computational time. Even though numbers of different classifiers have been proposed till now, still it is a very difficult problem for the researchers to model a correct classifier for superior diagnosis of different diseases. Therefore, in this work a new classifier called as kernel extreme learning machine (KELM) is proposed to classify the biomedical data. Again to enhance the classifier performance a optimization algorithm named as levy flight based fruit fly (LVFFO) is applied to optimized the parameters of KELM. The presented model performance is evaluated by comparing its performance with others models (SVM, ELM and ELM-FFO). The performances are evaluated by calculating some performance indices (G-mean, sensitivity, F-Score, accuracy, specificity, and precision). The results validate that KELM-LVFFO giving superior result among all the models.