Autism spectrum disorders were compared using EEG data, the Novel Gaussian Kernel Smoothing Classifier, and the KNN Classifier.

Abstract

Aim:The purpose of this research is to determine the causes of Autism Spectrum Disorder (ASD), the using ofcontemporary-day algorithms, and evaluating the accuracy and sensitivity rate between Novel Gaussian Kernel Smoothing and KNN.  Materials and Methods: The Datasets contain EEG signal value photographs from the hospital centers and are used in this research.The sample is interpreted as (N =40) for Novel Gaussian Kernel Smoothing and (N =40) for KNN, and the total sample is calculated using Clinical.com by maintaining the alpha error-threshold at 0.05, enrollment ratio as 0:1, 95%confidence interval, g-power at 80%. Accuracy and sensitivity was calculated by using standard dataset. Results: The level of accuracy and the level of sensitivity are compared using the independent IBM-SPSS sample testing software. There is a statistical indifference between the Novel Gaussian Kernel Smoothing algorithm and KNN. The accuracy of Novel Gaussian Kernel Smoothingalgorithm is 54.57% (p=0.001) is higher than KNN 50.4% and the sensitivity of Novel Gaussian Kernel Smoothing algorithm is 45.55% (P=0.001) is  higher than KNN 43.8%. Conclusion: Gaussian kernel smoothing algorithms appear to provide better accuracy and sensitivity than KNN in predicting autism with an EEG signal.