AdaBoost Ensemble Methods Using K-Fold Cross Validation for Survivability with the Early Detection of Heart Disease

Area of Science:

• Medical informatics
• Artificial intelligence in healthcare
• Cardiovascular disease research

Background:

• Large datasets for heart disease diagnosis present storage and processing challenges.
• Early and accurate diagnosis of heart disease is critical for effective medical treatment and patient survival.
• Heart disease is a growing global health concern, necessitating advanced diagnostic tools.

Purpose of the Study:

• To investigate the efficacy of machine learning (ML) techniques for early heart disease detection.
• To address data imbalance and noise issues in medical datasets using data mining preprocessing.
• To compare the performance of various ensemble classifiers for cardiac disease identification.

Main Methods:

• Implemented homogeneous and heterogeneous ensemble classifiers.
• Utilized Synthetic Minority Oversampling Technique (SMOTE) for data preprocessing to handle class imbalance and noise.
• Employed Naive Bayes (NB) and Decision Tree (DT) algorithms, along with their ensembles, for classification.

Main Results:

• The AdaBoost-Random Forest classifier demonstrated superior performance.
• Achieved an accuracy of 95.47% in the early detection of heart disease.
• SMOTE effectively mitigated data imbalance and noise, improving classifier performance.

Conclusions:

• Ensemble machine learning models, particularly AdaBoost-Random Forest, are highly effective for early heart disease detection.
• Data preprocessing techniques like SMOTE are crucial for optimizing ML model performance on imbalanced medical datasets.
• The proposed ML approach offers a promising tool for improving cardiovascular disease diagnosis and patient care.