Autistic Spectrum Disorder Screening: Prediction with Machine Learning Models

In present scenario autism spectrum disorder (ASD) is having faster momentum than ever before. Detecting autistic behavior and traits through screening test is time consuming, labor demanding and expensive. With growth of AI and ML, autism can be predicted in an early stage. Various studies have been carried out using different techniques, but these studies didn’t provide any proper conclusion about predicting autism traits in terms of different age groups. Therefore, this paper aims to propose an effective prediction model based on ML technique and to develop a user interface for predicting ASD for people of any age. As outcomes of this research, an autism prediction model was developed by merging Random Forest-CART (Classification and Regression Trees) and Random Forest-ID3 (Iterative Dichotomizer 3) and also a user interface was developed based on the proposed prediction model. The proposed model was evaluated with AQ10 dataset and 250 real datasets collected from people with and without autistic traits. The evaluation results showed that the proposed prediction model provides better results in terms of accuracy, specificity, sensitivity, precision and false positive rate(FPR) for the data.