Projects
Please use the links provided to access the GitHub source code and reports.
Evaluation on the Classification of Cardio-vascular Disease Using Logistic Regression
January 2024
An in-depth discussion of my implementation of Logistic Regression to classify patients with Cardiovascular Disease (CVD). This Python model got an accuracy of 73.07% due to the extremely restricted dataset given. The use of this model would be to aid medical professionals in diagnosis. Logistic Regression was chosen as it is easily comprehendible and lightweight enough to be widely distributed. Although there are some ethical concerns with using this as a classifier for CVD, the roots are there for utilising this on a more detailed dataset
Classification Of Systolic Heart Murmurs of
Children Using Dynamic Time Warping and
K-Nearest Neighbours
January 2024
An evaluation on the use of KNN and Dynamic Time--Warping for classification of Heart Murmurs in children. This method written in MatLab did not perform well and will be improved in the future. The dataset is incredibly noisy, although a real-life example of setting, making it a major issue for this classifier. The use of a primitive KNN is probably not the best approach to this challenge. The dataset was the dataset of the CirCor Moody PhysioNet Challenge 2022.