Predictive Analysis for daASD Using Population based Incremental Learning

Autism is a clinically defined behavioral syndrome that initially appears in childhood and reflects the underlying neurodevelopmental irregularities. The main aim is to define the clinical trajectories of dyslipidemia-associated ASD (daASD) is to take the massive data integration has revealed an emerging ASD subtype characterized by dyslipidemia. However, the clinical course of this subtype remains unknown. To analysis of Electronic Health Record (EHR), data identifying the pattern sequence of medical events experienced by people with daASD. Within these data, individuals with daASD will be identified as those having at least one ASD diagnosis and dyslipidemia diagnosis or repeated blood lipid test results that are outside the normal range. Comparisons will be made between individuals with daASD, ASD with no dyslipidemia (ASD+/dyslipidemia- ), dyslipidemia with no ASD (ASD-/dyslipidemia+), and individuals with no history of neuro-developmental disorders or abnormal lipid levels (ASD-/dyslipidemia-). Matching will be based on age, gender, ethnicity and address/zip code, as a proxy of socioeconomic status. The parents of each member of the above groups will be identified and their data will be collected. The direct comparisons of daASD with other ASD subtypes are required diagnosis for comorbid epilepsy/recurrent seizures and comorbid Fragile X syndrome using population based incremental learning. In this regard, Random Forest approach is on daASD for classification and semi-supervised learning is a process for daASD. The daASD experimental result is based on Electronic Health Records of Least Square Estimation Regression with Predict the Disease. © 2021 School of Science, IHU. All rights reserved.