Comparative Study on Detection of Heart Disease using KNN Algorithm

Nowadays, people are so occupied with earning money to buy things they aren't able to take any time for themselves, which can cause both physical and mental health problems. As a result, heart disease is widespread in modern society, especially in urban regions, where people experience higher rates of stress. Thus, in the so-called material world, heart disease has risen to prominence as a leading cause of death for both sexes. The disease has become the leading cause of death in both urban and rural areas. About 17 million individuals worldwide perish every year as a result of cardiovascular illness, most often manifesting as Heart Failure (HF). When the heart's blood pumping capacity is inadequate, HF results. Many lives could be prevented if heart disease could be predicted at an early stage. The primary goal of this study is to find ways around certain restrictions and to create a system that performs well and can reliably forecast the occurrence of HF. The Kaggle data collection, which includes 12 essential attributes, is used in this work. To further investigate the role of each attributes inducing HF, we conducted a comprehensive Exploratory Data Analysis. In this work, several K-Nearest Neighbour(KNN) algorithms, including classic KNN, adaptive KNN, and weighted KNN, are used. Results showed that adaptive KNN was the most effective method, with accuracies as high as 97.99%. In addition to describing how each method is implemented, the study provides a comparison of the various approaches using positive and negative metrics.