Comparison Model Optimal Machine Learning Model With Feature Extraction for Heart Attack Disease Classification
DOI:
https://doi.org/10.15294/sji.v11i2.4561Keywords:
Heart attack, Classification, Machine learning, PCA, ROC curveAbstract
Purpose: The purpose of this study is to classify the number of people affected by heart disease and those not affected by heart disease based on various categories of heart attack causes. This study aims to urge people to take better care of their health and to serve as a reference for doctors to educate patients about the dangers of heart attacks.
Methods: The model will be constructed via a machine learning methodology. The algorithms utilized in its development encompass the Support Vector Machine (SVM) algorithm, the K-Nearest Neighbor (k-NN) algorithm, and the Random Forest (RF) algorithm. This study utilizes principal component analysis (PCA) as a means of extracting optimized features from the dataset, employing techniques for dimension reduction prior to modeling the data.
Result: Cumulative explication of the concept of variance constitutes a foundational aspect of PCA (principal component analysis) within the scope of the current research, namely a dimensionality reduction technique employed in multivariate data analysis to facilitate model development, thereby enabling the creation of more optimal and comprehensive models. In this research, the dimensions of training data are incorporated during the process of model creation. The results show KNN model exhibits the highest performance, with an accuracy of 86%, precision of 86%, recall of 91%, and F1-score of 88%. Furthermore, evaluation using the ROC metric also provides a relatively favorable value, 0.85.
Novelty: Researchers used 1190 patient data sourced from Kaggle. Before modeling the algorithm, researchers conducted EDA & Preprocessing which includes missing values to find data that does not have information, then duplicate data to find duplicated data, there are 270 duplicated data, then the duplicated data is deleted so that the data becomes 737, then PCA implementation is carried out. PCA is reducing features automatically without changing the data.
