Early Detection of Autism Spectrum Disorder in Toddlers: A Fast and Efficient Machine Learning Approach

A patient affected by autism experiences significant impacts on their daily life, but early intervention can help mitigate its effects. In this research work, we introduced a more accurate and lightweight model to identify Autism Spectrum Disorder (ASD) using the Toddlers dataset. Several data processing techniques were applied to the dataset before prediction: SMOTETomek, independent t-tests, and MinMaxScaler for balancing the dataset, statistical analysis, and data normalization, respectively. Finally, we applied eight machine learning (ML) classifiers Decision Tree (DT), Catboost, XGBoost, Voting, Stacking, Extra Trees, Gaussian Naive Bayes, and Bernoulli Naive Bayes for classification after selecting the one most significant feature based on Chi-squared scores. The experimental results show that the DT classifier outperformed the other classifiers and the current state-of-the-art models, achieving an accuracy of 100% in early ASD detection. This research work provides insightful information for healthcare professionals for automating ASD screening.