Abstract:
"
Heart disease is one of the most common diseases affecting most people in their middle
or old age, which can eventually lead to fatal complications. A heart disease occurs
when the heart is unable to pump enough blood to all areas of the body. Heart condition
diagnosis must be accurate and timely in order to prevent and cure heart disease. Heart
disease is responsible for one third of all deaths worldwide. Every year, about 17
million people die from heart diseases, and the cardiovascular disease (CAD) is
especially prevalent in Asia due to their lifestyle. Non-invasive approaches such as
ensemble learning, machine learning, deep learning, are effective and efficient for
predicting healthy people and heart patients.
Wellbe system provide a service that enables medical personal and patients to check
whether that person is suffering from a heart disease. Heart disease perdition model
required 13 factors to identify heart disease. Ensemble learning enables to get detection
from multiple models. Proposed system uses KNN, Random Forest, Gaussian NB and
SVC models to develop the hybrid approach to identify heart disease.
Final model has the combination of all the models and also it is used by medical
personal and general public. To interact with the model web application is developed.
Final model has an accuracy of 85%. Wellbe is evaluated by end users, domain experts
and industrial experts.
"
