Attention Module-Based Fused Deep Cnn for Learning Disabilities Identification Using EEG Signal

Learning disabilities (LDs) are analyzed in children whose educational capabilities of understanding, inscription, or arithmetic are harmed as well as lagging under their age, schooling, as well as cleverness, which have a predictable occurrence among the percentage from 5 to 9 in the pediatric population. Preceding research regarding electroencephalography (EEG) signal havestated a delay in the growth of alpha-band at precise phenotypes of LD, which appears to provide a feasible clarification for differences in the maturation of EEG. Thus, EEG signals of children with reading disorders (RDs) are depicted through an advanced theta as well as a lesser alpha than those of characteristically raising children. Thus, an attention module based fused Deep CNN is developed for identifying learning disabilities using EEG signals. The main point of the proposed research is to predict the learning disabilities of children depending on EEG. The accuracy of the proposed method attains the values of 97.60% and 95.12% in terms of training percentage as well as k-fold.