Deep Learning Models for Early Identification of Learning Disorders in Children with Autism Spectrum Disorder

• Published in: SN computer science Vol. 6; no. 7; p. 761
• Main Authors: Kumar, H. S. Ranjan, Preethi, S., Fathima, Nasreen, Yuvaraj, B. K., Kumar, K. L. Santhosh, Bharath, K. N., Yathiraj, G. R.
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.10.2025; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Attention deficit hyperactivity disorder; Autism; Children; Children & youth; Citrus fruits; Communication; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Data Structures and Information Theory; Datasets; Decision trees; Deep learning; Disorders; Dyslexia; Explainable artificial intelligence; Identification; Information Systems and Communication Service; Intervention; Machine learning; Neural networks; Original Research; Pattern Recognition and Graphics; Software Engineering/Programming and Operating Systems; Toddlers; Vision; Children’s; Accuracy; Machine learning models; Autism spectrum disorder (ASD)
• ISSN: 2661-8907; 2662-995X; 2661-8907
• DOI: 10.1007/s42979-025-04308-3

Learning disorders are common among children with Autism spectrum disorder (ASD). Although autism itself is not a learning disability, it can significantly affect a child’s ability to process and retain information. And as a result, it hinders their academic and social progress. Early diagnosis of children who may have chances of developing learning disorder often helps to provide effective treatment. This research aims to develop mechanisms to detect and predict the learning disorder in children with ASD traits aged from 1 to 18 years old. In this work, the dataset has been obtained from Kaggle, which consisted primarily of 1985 different set of values. After data preprocessing, we obtained our final dataset with 1937 values. In total, nine machine learning algorithms were employed to predict whether a child with ASD traits has a probability of developing learning disorder or not. Two hyperparameter optimizers were employed to improve the predictability. Accuracies of 99.48% were obtained by both decision tree and random tree classifier. Finally, LIME, an explainable AI framework, was applied to interpret and retrace the prediction output of the machine learning models.