Effect of Feature Selection to Improve Accuracy and Decrease Execution Time with Predicating Learning Disabilities in School Going Children

Learning disability in school children is the representation of brain disorder which includes several disorders in which school going child faces the difficulties. The evaluation of learning disability is a crucial and important task in the field of educational field. This process can be accomplished by using data mining approaches. The efficiency of this approach is based on the feature selection while performing the prediction of the learning disabilities. In paper mainly aims on the efficient method of feature selection to improve the accuracy of prediction and classification in school going children. Feature selection is a process to collect the small subset of the features from huge dataset. A commonly used approach in feature selection is ranking the individual features according to some criteria and then search for an optimal feature subset based on evaluation criterion to test the optimality. In the Wrapper model we use some predetermined learning algorithm to find out the relevant features and test them. It requires more computations, so if there are large numbers of features we prefer to filter. In this paper first we have used feature selection attribute algorithms Chi-square. Info Gain, and Gain Ratio to predict the relevant features. Then we have applied fast correlation base filter algorithm on given features. Later classification is done using KNN and SVM. Results showed reduction in computational cost and time and increase in predictive accuracy for the student model. The objective of this work is to predict the presence of Learning Disability (LD) in school-aged children more accurately and help them to develop a bright future according to his choice by predicting the success at the earliest.