Early Detection of Autism in Toddlers Using Machine Learning

This project aims to assess autism risk in young children using machine learning based on questionnaire data. Key questions are selected using the Lasso technique, and an ensemble model combining XGBoost, Random Forest, and Gradient Boosting achieves 94.79% accuracy. The tool is designed to help parents and caregivers with early screening and awareness of Autism Spectrum Disorder. It was developed by CSE students Ramya, Sharanya, Shreya Shetty, and Vaishnavi Bijoor under the guidance of Prof. Savita Shenoy.