Autism classification using machine learning techniques / Asmaa Ahmed Elsayed Ahmed ; Supervised Hesham Ahmed Hefny , Nagwa A. Meguid , Mahmood A. Mahmood

Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Computer and Information Sciences

Autism is a neurodevelopment disorder which is characterized by weak social interaction disorder, weak verbal communication and non-verbal, behavioral patterns of restricted and repetitive. The impact of autism on the natural growth of the brain in the area of social life and communication skills, where children and people usually face with autism difficulties in the field of non-verbal communication, social interaction, as well as difficulties in recreational activities. Many of the available diagnostic tools are commonly used in autism researches such as the diagnosis of Autism Interview Revised, which is defined as semi-structured interview done with the parents and childhood autism rating scale. Both are used to diagnose autism and determine the degree of autism: medium, heavy. The proposed research model compiles data on cases to diagnose autism to determine its degree. The data has been collected regarding the autism cases at 2the National research center in Egypt3. The proposed research model consists of three phases: pre-processing data, single label classification, multi label classification. Pre-processing data: Divided into two parts3 data collation and data processing2, data were available in paper form and were collected and incorporated on a computer to use it, after data entry is going to process the data and in this part have been limited to the particular problems with data like data imbalance

Issued also as CD