Early Detection, Classification and Support in Learning Disabilities Using Machine Learning Techniques

Early diagnosis of learning disability in forms of dyslexia, dyscalculia, and attentiondeficit/hyperactivity disorder (ADHD) is important in enhancing academic achievement and psychological wellbeing of children. The traditional methods of diagnosis are however cumbersome, costly, and unprovisioned. The present paper demonstrates a smart web-based screening and assistance unit that incorporates machine learning to identify and classify the presence of learning disabilities in their initial stages with the help of an interactive WebApp built with Streamlit. The system is proposed and combines several assessment modules such as imagebased cognitive quizzes, behavioral surveys, sentence-based reading and writing analysis, and mental disorder prediction model. The reaction of the user is converted into domain specific characteristics like language proficiency, memory, visual discrimination, auditory perception, and processing speed which is later processed through trained machine-learning classifiers. Besides this, a chatbot powered by AI is also integrated to offer real-time instructions and accessibility facilitation.