Date of Publication :23rd February 2018

Abstract: Presently there has been good interest in computing similarity for data mining and machine learning particularly. In this paper, we are discussing Kernel methods. Kernel methods are used pervasively mainly because of its large number of applications and scientific challenges. It has the capability to model real-world data and give efficient solutions to real-world problems. Such solutions given are one of its kind the most accurate and efficient as compared to the other existing ones. Along with the popular applications of kernel this paper mainly gives a basic understanding of fundamental aspects of kernel methods their underlying principles and formulas used.Various aspects of kernel approach are discussed.Some simulation results are shown in the discussed kernel methods and compared with new proposed kernel design.

Reference :