Premature Identification of Autism Spectrum Disorder using Machine Learning Techniques

Abstract

Autism Spectrum Disorder (ASD) is gaining traction quicker than ever before. Autism features can be detected by screening tests, but they are costly and time consuming. Autism can now be predicted within 12 to 36 months thanks to advances in artificial intelligence and machine learning (ML). The suggested model will be tested using the AQ-10 dataset as well as 1000 real-world data from people with and without autistic symptoms. These datasets contain a collection of questions with only two possible answers: yes or no. Autistic children's parents will have noticed a specific pattern in their behavior, and they will be able to answer the questions based on these observed behavioral patterns. After collecting all the inputs in the form of excel sheet further it will be fed into the proposed system. With the usage of different algorithms such as Random Forest, Support Vector Machine and Ada boost the system will process the data efficiently and give the output. For both types of datasets, the analyzed findings demonstrate that the proposed prediction model will deliver improved outcomes in terms of accuracy, specificity, precision, and false positive rate (FPR).