Personalized ASD Interventions Recommendation for Sri Lankan Patients

Autism Spectrum Disorder (ASD) is a complex brain development disorder that impacts how an individual understands and socializes with others, resulting in social, communicational, and behavioral issues. Treatment effectiveness varies, hindering data-driven decisions. Sri Lanka lacks specialized facilities, necessitating culturally attuned methods. We propose a solution: employing machine learning for cost-effective, personalized interventions. Collaborative and Content-based filtering recommend tailored interventions. A Support Vector Machine predicts culturally sensitive treatment responses. The models are trained and evaluated using a dataset of medical metadata. The Mean Squared Error (MSE) of CF, CBF, and SVM, models are 0.584, 0.029, and 0.0897 respectively. The SVM model outperforms Random Forest (RF) models and Decision Tree (DT) models in predicting multi-regression outputs. The results show that the use of a combination of CF, CBF, and SVM models can accurately recommend personalized interventions for ASD patients. The findings of the research are culturally sensitive to Sri Lanka and can be used to recommend personalized interventions which have the highest effectiveness for a patient.