Identification of heart problems through predictive models based on supervised machine learning
Keywords:
XGBoost, Random Forest, KNeighbors Classifier, Health, HeartAbstract
The predictive models of "Supervised machine learning" are becoming increasingly important in aiding decision-making for various areas of human knowledge and, consequently, will also be important for assisting in medical cases of greater complexity. The goal of the study is to develop a supervised machine learning algorithm that can have a high success rate in predicting whether a person has a heart problem or not. The article shows how the models were developed, the tests applied before the implementation of the models, the utilization rate of each model and an analysis of which is the most efficient model for a specific situation. The specifications of each supervised machine learning model and its impact on the development of the models that were used in the work, determined by testing and applications made in the Python programming language; The positive and negative results were considered to reach a final position on what was the best way to use the algorithms in this case. The article concludes that the application of supervised machine learning models in the diagnosis of heart problems can help many health institutions, both public and private, to streamline processes and increase the success rate when classifying a person as cardiac or non-cardiac, to have a high improvement in this process and consequently increase efficiency and profitability in the case of private institutions.
