| Resumo: | Machine learning, broadly speaking, applies statistical methods to training data to automatically adjust the parameters of a model, rather than a programmer needing to set them manually. Deep Learning is a sub-area of Machine Learning that studies how to solve complex and intuitive problems. The methodologies adopted, using computational means, such as the machines learned and those understood in the world in specific contexts from previous experiences and based on the hierarchy of concepts, use the most used concepts for the form and efficient solution of more varied complex problems. The main objective in this work is to study various classification algorithms in the area of machine learning, and validate until these points can use a solution for choosing more accurate methods in the selection of tests and in new statistics to improve the therapeutic response. The data involved in the training of classification algorithms refer to all patients with metabolic diseases shredding between the years 2003-2021 and the retrospective part. The best classification algorithms to develop are used in the decision support system in the most effective way in choosing the appropriate therapy for each of the future patients who predicted an approximate rate of 20 patients per year.
|
|---|