CoChat

Effective Heart Disease Prediction Using Hybrid Machine Learning Techniques

Open access

The DOI matches its Crossref record; 47 of the 51 references it cites were found in a registry or an index.

Abstract

Heart disease is one of the most significant causes of mortality in the world today. Prediction of cardiovascular disease is a critical challenge in the area of clinical data analysis. Machine learning (ML) has been shown to be effective in assisting in making decisions and predictions from the large quantity of data produced by the healthcare industry. We have also seen ML techniques being used in recent developments in different areas of the Internet of Things (IoT). Various studies give only a glimpse into predicting heart disease with ML techniques. In this paper, we propose a novel method that aims at finding significant features by applying machine learning techniques resulting in improving the accuracy in the prediction of cardiovascular disease. The prediction model is introduced with different combinations of features and several known classification techniques. We produce an enhanced performance level with an accuracy level of 88.7% through the prediction model for heart disease with the hybrid random forest with a linear model (HRFLM).

Overview

CoChat Summary

TL;DR

The paper proposes HRFLM, a hybrid model combining Random Forest and a linear model, to predict heart disease from the UCI Cleveland dataset. It reports improved accuracy over several individual classifiers, with the abstract citing 88.7% and Table 3 listing 88.4% for HRFLM.

Key findings

  • HRFLM (proposed) achieved accuracy 88.4%, classification error 11.6%, precision 90.1%, F-measure 90%, sensitivity 92.8%, and specificity 82.6% in Table 3. The abstract states an accuracy level of 88.7%.
  • In the same comparison, other models included Naive Bayes (75.8% accuracy), Generalized Linear Model (85.1%), Logistic Regression (82.9%), Deep Learning (87.4%), Decision Tree (85%), Random Forest (86.1%), Gradient Boosted Trees (78.3%), Support Vector Machine (86.1%), and VOTE (87.41%).
  • The dataset contained 303 patient records; 6 records with missing values were removed, leaving 297 patient records. After pre-processing, 137 records indicated presence of heart disease and 160 indicated absence.
  • The proposed hybrid combines Random Forest and Linear Method (LM); the paper reports that RF and LM were the best-performing individual methods in its rule-based dataset splits.

Methods

  • Data source: UCI machine learning repository, Cleveland database, selected because it is commonly used and has comprehensive records. The Cleveland dataset has 76 attributes, but the repository provides a subset of 14 attributes; 13 attributes were used for prediction.
  • Pre-processing and modeling: R Studio Rattle was used. The ML process included data pre-processing, feature selection based on Decision Tree entropy, classification modeling, performance evaluation, and repeated feature/model combinations. The experiment used NB, GLM, LR, DL, DT, RF, GBT, and SVM with all 13 attributes.
  • Evaluation: models were assessed with a confusion matrix (TP, TN, FP, FN) and metrics including accuracy, precision, classification error, F-measure, sensitivity, and specificity. One section states 252 records were used for training and the remaining for testing.

Limitations

  • The paper states that further extension is desirable to direct investigations to real-world datasets instead of just theoretical approaches and simulations, and that new feature-selection methods can be developed to better understand significant features.
  • Evident from the text: the study uses only the Cleveland dataset and a small sample of 297 records, with no external validation reported.
  • Evident from the text: reported accuracy is inconsistent between the abstract (88.7%) and Table 3 (88.4%), and several result tables are fragmented or difficult to interpret.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1109/access.2019.2923707

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 4 Not valid · 0 Not checked · 43 Own results and interpretations

29 more claims

Create a free account to see all 47 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 11
  • It may state two separate claims as one. 4
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 47 references

Research works printed but not linked: 3

  1. wagh, 2016, CDSS for heart disease prediction using risk factors, International Journal of Research & Innovation in Computer Engineering, 4, 12082
  2. raju, 2016, Mining techniques, Proc Conf Emerg Devices Smart Syst (ICEDSS), 253
  3. al-milli, 2013, Backpropogation neural network for prediction of heart disease, J Theor Appl Inf Technol, 56, 131

Cited by

Items 1,901–1,954 of 1,954

Metadata

Authors
Mohan, S., Thirumalai, C., & Srivastava, G.
Affiliations
  • Vellore Institute of Technology University (IN)
  • Brandon University (CA)
Venue
IEEE Access
Published
2019
Volume
7
Pages
81542-81554
ISSN
2169-3536 (ISSN-L)
OpenAlex
W2949767632
Cited by
1,951
References
51
Citations by year
2026: 179 · 2025: 308 · 2024: 372 · 2023: 442 · 2022: 334 · 2021: 219 · 2020: 88 · 2019: 8
FWCI
176.85
Type
journal article
Language
en
Open access
Locations
  • IEEE Access (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
China Medical University
Topics
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
  • Physical Sciences › Computer Science › Information Systems › Data Mining and Machine Learning Applications
  • Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › ECG Monitoring and Analysis
Keywords
heart disease prediction, cardiovascular disease, hybrid random forest, linear models, feature selection, medical data analysis, classification accuracy, Internet of Medical Things

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.