Kernel-Induced Matriarch Path Tracking Elephant Herding Optimization Technique for Identification and Classification of Cancer Types Using Support Vector Machines

Abstract

Abnormal cells in the human body that keep on mutating are termed to be cancer in medical terms. There are multiple types of cancer identified in human beings. It is very much essential to identify and classify the type of cancer in its earlier stage. This objective can be satisfied by artificial intelligence which has a subfield of machine learning to create a generalized model that could identify and classify cancer with increased performance. To perform the identification and classification of various cancer types, in this paper, two techniques are adopted. The optimized feature set computation was done using the Kernel-Induced Matriarch path tracking Elephant Herding Optimization (KIM-EHO) and the classification for the given samples was done using the Support Vector Machines (SVM). The proposed techniques are implemented with the benchmark datasets and the results proved that the proposed methodologies outperformed the existing methods in terms of accuracy, specificity, sensitivity and time complexity.