With the recent prevalence of machine learning and data mining, significant effort is being made to push the frontier by which computers can assist humans in deriving insights and in making decisions through the analysis of a multitude of increasingly complex as well as heterogeneous datasets. Support Vector Machines (SVMs) is a type of data driven machine learning approach that deals with predictive classification. Using a large set of observations with known labels (training set), SVM finds a maximum margin function that separates the observations into two or more classes where each observation is a point in a multidimensional space of feature measurements. New unlabeled data are then assigned a class based on their geometric position relative to the classifier function. Given the vast amount of complex features that modern systems use, finding the classifier function often requires the simplification of the features space by identifying the dimensions that have the most distinguishing power. It is therefore essential to jointly optimize the feature selection and the classification in order to ensure the best performance. In this part of research work, Routine Correspondence Method with Grey Wolf Optimization based Imperforate Support Vector Machine Classifier (ISVMC) for High Dimensional Datasets is presented.
Volume 11 | 01-Special Issue
Pages: 652-660