High Dimensional Datasets Optimization handling by Wrapper Sequential Feature Selection in Forward Mode - A Comparative Survey

Abstract

High-quality data might be difficult to be produced when there is a large quantity of information in a single educational dataset. Researchers in the field of educational data mining have recently begun to rely more and more on data mining methodologies in their investigations. However, instead of undertaking feature selection methods, many research investigations have focused on picking appropriate learning algorithms. Since these datasets are computationally complicated, they need a lot of computing time for categorization. This article examines the use of wrapper approaches for the purpose of managing high-dimensional datasets in order to pick appropriate features for a machine learning approach. This study then suggests a strategy for improving the quality of student or educational datasets. For future investigations, the suggested framework that utilizes filter and wrapper-based approaches may be used for many medical and industrial datasets.