Diagnosing heart problem using machine learning
Authors
Juberkhan, Pathan
Issue Date
2020
Degree
MSc in Data Analytics
Publisher
Dublin Business School
Rights holder
Rights
Items in eSource are protected by copyright. Previously published items are made available in accordance with the copyright policy of the publisher/copyright holder.
Abstract
The main part of the metabolism in the case of the human body is the requirement of the oxygenated blood so that the process can be complete. In the result of the metabolism, the wastes and the deoxygenated blood needs to be pumped out from the different organs present in the body so that the human body can sustain the process of living. Heart being the pumping organ of the body, not only supplies the oxygenated blood to the different parts of the body but also removes the wastes and the deoxygenated blood. Hence, to take the proper care of the heart the regular check up is very important. There can be many reasons which can be due to genetic history and some of the acquired habits that can result in an adverse effect to the heart. There has been a subsequent amount of research works that takes into the account the prediction of the well being of the human heart. In this research we have analysed around 1095 patient cases and tried to identify the important risk factors that can be a primary reason for the heart problems. This research work tries to present a work that can be an immediate step to find a probability score for the heart problem. The different risks factors are those which can be the primary reason for the occurrence of the heart problem. In this research we have analysed different classification models to identify the heart problem. The data collected is from five different cities in India and from different age groups. The primary aim of such activity is to present a transparent solution which can help the patient to know statistically whether there is any probability of happening of any heart problem. This solution is in fact not a replacement for a medical practitioner but rather aiding some help in the diagnostic process of any doctor. This in order creates a transparency in the treatment between a doctor and a patient. The model is checked with the number of False Positives and the False Negative that the model performed and based on that the best algorithm is selected for the prediction.