Classification of Heart Failure Using Machine Learning: A Comparative Study.

Area of Science:

• Cardiology
• Data Science
• Machine Learning

Background:

• Heart failure is a significant global health concern requiring accurate diagnostic tools.
• Machine learning offers promising approaches for improving the early detection and management of heart failure.

Purpose of the Study:

• To evaluate and compare the performance of various machine learning classification algorithms for heart failure prediction.
• To identify the most effective machine learning model for accurately diagnosing heart failure cases.

Main Methods:

• A dataset of heart failure indicators was used to train and test multiple classification algorithms.
• Algorithms evaluated included logistic regression, random forest, decision tree, K-nearest neighbors, and multilayer perceptron (MLP).
• Performance metrics such as specificity, Area Under the Curve (AUC), and Matthews Correlation Coefficient (MCC) were used for comparison.

Main Results:

• The random forest model achieved superior performance with specificity = 0.93, AUC = 0.97, and MCC = 0.83.
• Random forest demonstrated high accuracy, establishing it as the best-performing model in this study.
• K-nearest neighbors and multilayer perceptron (MLP) exhibited lower accuracy rates compared to random forest.

Conclusions:

• The random forest algorithm is highly effective for identifying heart failure cases.
• The study emphasizes the critical role of feature selection, data quality, model choice, and hyperparameter tuning in machine learning for healthcare.
• Machine learning techniques are valuable tools for advancing heart failure diagnosis and management.