MACHINE LEARNING ALGORITHM BASED PREDICTION OF HEART STROKE CHANCES BASED ON DATA PROVIDED

The main goal of this paper is to predict the chances of Heart strokes by using data which is already available. The data set is divided into 2 parts, The first part of the data set consists of 67% of data and is used to train models. The second part of data which is 33% of data is used to test the model. The classifier model is prepared using machine learning algorithms like Decision Trees and predicting the chances of Heart stroke based on various factors like age, chest pain, blood pressure, cholesterol levels, blood sugar levels and maximum heart rate achieved. The fields of the data set mostly represent and have the values which highly contribute for the heart stroke possibilities. The data set has to be cleaned by using various data cleaning techniques and removing null or incorrect values from the dataset after the data set has been cleaned. The data set is provided to machine learning algorithms which classify the data. Similarly we can also diagnose several diseases by collecting the data based on the primary causes of diseases and predicting the possibilities of these diseases in patients. By predicting the possibilities of heart stroke by using basic medical test data we can save the time as well as money on highly costly diagnosis.