A Multivariate Analysis using Mathematically Modified Machine Learning (MA- MM-ML) Algorithm of Autism Spectrum Disorder (ASD)

Abstract

  Autism Spectrum Disorder (ASD) is a neuro-developmental disorder which impacts a person’s ability to interact with others, communicate and engage in activities in a repetitive manner. Thus, the purpose of this study is to develop a Multivariate Analysis for the diagnosis of ASD in children through Mathematically Modified Machine Learning (MA-MM-ML) algorithm to enhance the efficacy and efficiency of the diagnostic process. Standard diagnostic procedures are time-consuming and approximate, thus, the use of advanced computational methods is necessary. Therefore, the reason for this research is to apply multivariate analysis and machine learning algorithms that can detect the heterogeneity of the ASD and diagnose the disease whereas develop interventions at an early stage. The MA-MM-ML algorithm that has been proposed contains the preprocessing step, the PCA, and BO-SVM techniques with hyperparameter tuning. This approach ensures that the ASD analysis and classification are accurate and less resource intensive thus improving the diagnostic capability. Therefore, the present study indicates that the research work can be done with multivariate analysis and machine learning to overcome the challenges related to ASD research and practice. The experimental analysis of proposed MA-MM- ML is compared with existing state-of-art techniques whereas MA-MM-ML achieves highest accuracy of 92.67%.