ANÁLISE DE CARACTERÍSTICAS A PARTIR DE ALGORITMOS DE APRENDIZAGEM DE MÁQUINA PARA AUXÍLIO AO DIAGNÓSTICO DO TRANSTORNO DO ESPECTRO AUTISTA

Resumo

Machine learning algorithms are being successfully appliedin many areas of knowledge. In digital health, these onespermit for professionals to be helped to diagnose diseasesand disorders in advance and more accurately, contributingto the eective treatment of their patients. In this context,the proposed methodology makes use of a public databaseon Autistic Spectrum Disorder and the most relevant characteristicspresented by a patient. So, it analyzes with theDecision Tree, Support Vector Machine, Multi-Layer Perceptronand K-nearest neighbor algorithms to the constructionof a model capable of simplifying a decision strategy tohelp the diagnosis of this disorder type.