Abstract: Neurodevelopmental disorders (NDDs) usually develop during early childhood and affect a person’s ability to think, feel, and interact with others. Common conditions associated with NDDs include ASD, ADHD, as well as other issues like intellectual disabilities, learning challenges, and cerebral palsy. This study introduces a machine learning method to classify different types of NDDs into multiple categories using both traditional and deep learning models. A dataset containing more than 6,000 entries was created, including information about gender, age, and clinical symptoms. The data was cleaned and prepared using label encoding and feature scaling methods. Two models were built and tested: a Support Vector Classifier (SVC) with a linear kernel and a DNN that uses a fully connected structure based on CNNs. The SVC model achieved an accuracy of 84%, while the deep learning model performed better, with an accuracy of 88%. This was verified using ROC curves and AUC analysis. These findings demonstrate that deep learning is effective in identifying complex patterns in data and may help enhance diagnostic processes for individuals with neurodevelopmental disorders.

Keywords: svc, linear svc, machine learning, mental health, neurodevelopmental, CNN.

Downloads: | DOI: 10.17148/IARJSET.2025.12932