A Multi Layer Ensemble Learning Framework for Learning Disability Detection in School-Aged Children

The aim of the paper is to exhibit the importance of ensemble techniques in the prediction of learning disabilities in school-age children. Learning disability (LD) is a neurological disorder in which students have persistent difficulty in learning. About 1 out of 10 school age children in India suffer from learning disorders. Children with problems in LD are of a higher concern to the parents and teachers. Therefore, diagnosis of LD at the early stages play a major role. This study helps to identify the important parameters of LD thereby estimating the relevance of each symptom of LD and effectively predict the possibility of LD using ensemble technique. In this paper, stacking ensemble technique is implemented to predict whether a child is suffering from learning disabilities. Initially, the results of individual machine learning models that is logistic regression, k-nearest neighbour (KNN) and support vector machine (SVM) are calculated. Later, the above algorithms are used in stacking ensemble model along with decision tree and final results are obtained. Finally, the results of individual machine learning algorithms and stacking ensemble algorithm are compared.