A minor project implemented as part of my Engineering degree curriculum.
Using the Cleaveland Database of UCI repository for Heart Disease detection and the Breast Cancer Wisconsin Diagnostic dataset, a model has been defined using KNN to determine which patient in a pool of patients, with their medical characteristics in consideration, is most likely to be diagnosed with a heart disease or breast cancer depending on the dataset being analysed.
Post data cleaning, standardization and parameter tuning, The model's accuracy comes up to 88.52% for heart disease detection and 97.3% for breast cancer detection.