Abstract:
Human life is the most important asset of human beings. Every year millions of people lose their lives to cardiovascular diseases. It is a group of diseases related to blood vessels and the heart. Chances of developing cardiovascular diseases in a person can be controlled by reducing some risk factors that cause them. If they are predicted timely in patients, the patients can take decisions and make changes to their lifestyles, and consequently reduce the risk of developing cardiovascular diseases. The traditional way to predict cardiovascular diseases is to consult a medical specialist. But this method can be inaccurate, consumes a lot of time and is expensive. In the past, work has been done regarding predicting cardiovascular diseases in patients using machine learning and deep learning techniques. But there was a gap in the literature that was filled by this study.
In the proposed research, a model consisting of five main modules; data collection, data preprocessing, data processing, performance layer, and validation has been used. The proposed model gave very promising results. It has proven to be very efficient in predicting cardiovascular disease in a person using Gradient Boosting Tree algorithm. The model had 78.78%, 76.78%, 81.10%, 82.43%, 18.90% accuracy, sensitivity, specificity, miss rate, and precision, respectively. Moreover, the fallout, LR+, LR-, and NPV were 18.90%, 4.06, 3.82, and 75.16% respectively. The classification time was a few milliseconds per record and the detection time was approximately 0.2137 seconds per record. The proposed model also outperformed various well-known machine learning algorithms and state-of-the-art models.
