NDeep Learning Heart Stroke Prediction Model Integration of MMAM with NB(MMAM-NB) and DT(MMAM-DT)

A framework proposed which works by developing different initial centroid selection method called Maximum and Minimum Attribute Method (MMAM) for K-means clustering, integrated with various classification algorithms like Naïve Bayes and Decision Tree, res ulted in two models, model-1(MMAM-NB) and mo...

Full description

Saved in:
Bibliographic Details
Published in2023 IEEE 5th International Conference on Cybernetics, Cognition and Machine Learning Applications (ICCCMLA) pp. 373 - 380
Main Authors Priyadarshini, T. Swathi, Hameed, Mohd Abdul, Qadeer, Shadan Amatul
Format Conference Proceeding
LanguageEnglish
Published IEEE 07.10.2023
Subjects
Accuracy
Area under ROC
Decision Tree
Deep learning
Health care systems
health-related survey
Heart
high risk
Initial Centroid Selection Method
k-means Clustering
Lead
low risk
Machine Learning
Medical services
Naive Bayes
Predictive models
risk factors
Sensitivity
Sensitivity and specificity
Severity of Heart Stroke
Specificity
Online AccessGet full text

Cover

More Information
Summary:A framework proposed which works by developing different initial centroid selection method called Maximum and Minimum Attribute Method (MMAM) for K-means clustering, integrated with various classification algorithms like Naïve Bayes and Decision Tree, res ulted in two models, model-1(MMAM-NB) and model-2(MMAM-DT), and examine the impact of clustering technique has on classification algorithms, and a way for extracting the most important risk factors causing the severity of heart stroke for the purpose of perfect prediction during classification. We Calculated sensitivity, specificity and accuracy values and assessed comparison of two. As a final assessment, we plotted ROC curve and estimated AUC-score of two models. We conclude that the MMAM derivative method demonstrated a massive improvement in AUC- ROC scores when compared to previous studies, in comparison with the traditional random selection method of k-means. Also, MMAM-NB model outperformed in comparison with MMAM-DT, accuracy of 90%, AUC-ROC score of 0.96, 89% Sensitivity and 92% Specificity. Thus, integration of machine learning and deep learning in health care fields, results in massive improvement of the quality of patient's health.
DOI:10.1109/ICCCMLA58983.2023.10346833