Project information
- Category: Machine Learning
- Type: Binary Classification, Functional PCA, B-splines
- Client/Purpose: This experiment was intended to compare the B-splines and FPCA algorithms on a ECG data set
- Project date: August, 2020
- Project URL: github
Utilizing B-splines and Functional PCA for classification of abnormal ECGs
In medicine, an electrocardiogram (ECG) is an exam that allows physicians to detect a heart disease. In recent years, there has been an increasing interest in using computers to detect cardiac abnormalities. The aim of this problem is to achieve recognition of abnormal ECG results, by treating the ECG signals as functional data, extracting relevant features and then using a classification method to discriminate between normal and abnormal ECGs.
For the experiment, both B-splines and Functional PCA were used and compared after using a classifier (Random Forest). Each curve has 96 observations, that were reduced to 10 observations with B-splines with 8 knots. For FPCA, 10 functional principal components are calculated as well. The resulting confusion matrices and accuracy metrics indicate that B-splines was a better choice for this problem.