Interpretable Random Forest-Based Heart Disease Prediction with Clinical Feature Importance Analysis

Q: What is the clinical evidence from this study?

Population: Heart disease (n=1025). Intervention: Random Forest classifier (100 decision trees, Gini impurity criterion). Primary outcome: Binary heart disease classification accuracy and ROC-AUC on independent 20% testing subset.

Q: What does this research mean for the field?

A Random Forest classifier achieves 92.7% accuracy and a ROC-AUC of 0.94 in binary heart disease prediction using 13 clinical and demographic features. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

K Sasikala; Mahavishnu; Bharanidharan P.; Gowtham S.; Devanesh R.; Ramkumar D.

doi:10.36548/jeea.2026.007

October 28, 2026Journal of Soft Computing ParadigmOpen Access

Interpretable Random Forest-Based Heart Disease Prediction with Clinical Feature Importance Analysis

View Full Paper

Ask AI

Bookmark

Share

Key result

Random Forest classifier achieves ~93% accuracy for heart disease prediction.

n=1,025

Why the study?

Heart disease prediction from heterogeneous clinical attributes requires ML models capable of capturing nonlinear relationships among physiological and diagnostic variables.

Does a Random Forest-based machine learning classifier accurately predict binary heart disease status from 13 routine clinical and physiological features?

Key Points

To develop an interpretable Random Forest machine learning framework to classify binary heart disease risk and quantify the importance of clinical features.
Utilized the public Heart Disease dataset comprising 1,025 samples and 13 input clinical and demographic features.
Partitioned data into an 80-20 training-validation split to train a Random Forest classifier composed of 100 decision trees utilizing the Gini impurity index.
Evaluated performance on an independent testing subset of approximately 205 cases using accuracy, precision, recall, F1-score, ROC-AUC, and feature importance analyses.
The Random Forest model achieved an accuracy of 92.7%, precision of 92.6%, recall of 91.9%, F1-score of 92.2%, and a ROC-AUC of 0.94 on the testing subset.
Feature importance evaluation identified key predictive factors including chest pain type, thalassemia, maximum heart rate, ST slope, number of major vessels, age, blood pressure, and cholesterol.

Structured PICO

Does a Random Forest-based machine learning classifier accurately predict binary heart disease status from 13 routine clinical and physiological features?

P

Population

1,025 patient records with 13 clinical, physiological, and exercise-related input features from the publicly available Kaggle Heart Disease dataset evaluated for binary heart disease classification.

E

Exposure

Random Forest (RF) machine learning classification model trained with 100 decision trees and Gini impurity splitting criterion, incorporating data pre-processing, categorical feature encoding, class probability estimation, and clinical feature-importance analysis.

O

Outcome

Binary heart disease classification performance measured by accuracy, precision, recall, F1-score, ROC-AUC, and confusion matrix analysis on the held-out testing subset.

A Random Forest classifier utilizing 13 clinical attributes achieved 92.7% accuracy and a 0.94 ROC-AUC in predicting binary heart disease on a public dataset while providing feature importance rankings for model interpretability.

Limitations

Application of this framework outside of the selected dataset requires further validation
Requires evaluation using large clinical datasets and generalization to different patient populations
Lack of comparison to other machine learning and deep learning models in the experimental setup
Needs evaluation using large clinical datasets and generalization to different patient populations
Lacks comparison to other machine learning and deep learning models in the current experimental setup

Implication

Should not yet change clinical prediction practices; leaves open prospective validation in real-world cohorts.

Abstract

Heart disease prediction from heterogeneous clinical attributes requires ML models capable of capturing nonlinear relationships among physiological and diagnostic variables. This research presents a Random Forest-based methodology for classification of binary heart disease data utilizing the publicly available Heart Disease dataset which includes 1,025 samples with 13 input features. The proposed approach includes data quality verification, categorical feature encoding, target generation, 80-20 training-validation split, ensemble classification, probability generation, and feature importance analysis. The RF model trained on 100 decision trees and Gini impurity index is measured for accuracy, precision, recall, F1-score, ROC-AUC, and confusion matrix analysis. This model achieves an accuracy of 92.7%, along with a ROC-AUC score of 0.94, demonstrating consistent classification performance on the independent testing subset. Analysis for feature importance reveals that some of the significant features include blood pressure, cholesterol, age, and heart rate. This model provides an explainable computational approach for heart disease risk classification.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Sasikala et al. (2026) studied Heart disease (n=1,025). Random Forest classifier (100 decision trees, Gini impurity criterion) was evaluated on Binary heart disease classification accuracy and ROC-AUC on independent 20% testing subset. The Random Forest classifier achieved 92.7% accuracy and a ROC-AUC of 0.94 for binary heart disease prediction on the independent testing subset.

synapsesocial.com/papers/6ae43436d9418cd7c8f47f72 https://doi.org/10.36548/jeea.2026.007

View Full Paper

Ask AI

Bookmark

Share

Interpretable Random Forest-Based Heart Disease Prediction with Clinical Feature Importance Analysis

Key Points

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Interpretable Random Forest-Based Heart Disease Prediction with Clinical Feature Importance Analysis

Key Points

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading