Assessing the Robustness of Machine Learning Algorithms for Cardiovascular Disease Detection Across Diverse Clinical Datasets DOI

Moulia Das Proma,

Khadiza Akther,

Bushra Siddika Priya

et al.

Published: Oct. 14, 2023

Cardiovascular diseases (CVD) continue to pose significant health risks globally, accentuating the need for early and precise detection mechanisms. With evolution of computational methods in healthcare, machine learning offers transformative solutions diagnostic accuracy. This research aims identify an algorithm with consistent performance across multiple datasets potential integration into a cardiac disease prediction platform. We examined nine prominent algorithms, namely Support Vector Machine (SVM), Gradient Boosting (GB), Random Forest (RF), Logis-tic Regression (LR), Decision Tree (DT), K-Nearest Neigh-bor (KNN), Naive Bayes (NB), Extreme (XGBoost), Multilayer Perceptron (MLP), evalu-ated their predictive two heterogeneous datasets. Both encompass 14 attributes but differ instance sizes: 303 1025, respectively. Through meticulous methodological framework, data underwent preprocessing, splitting, model training, followed by validation using metrics such as Precision, Recall, F1 score, Accuracy, coupled confusion matrix detailed class-based evaluation. Our findings revealed that MLP algorithms exhibited superior consistency robustness both datasets, achieving peak accuracy 95.14%. While XGBoost performed proficiently on one dataset, its wavered cross-dataset scenario. Based these findings, either or models are recommended developing robust heart system. not only affirms revolutionizing CVD diagnostics also underscores importance selection based dataset characteristics.

Language: Английский

Classification models combined with Boruta feature selection for heart disease prediction DOI Creative Commons

G. Manikandan,

B. Pragadeesh,

V. Manojkumar

et al.

Informatics in Medicine Unlocked, Journal Year: 2024, Volume and Issue: 44, P. 101442 - 101442

Published: Jan. 1, 2024

Cardiovascular disease (CVD), generally called heart illness, is a collective term for various ailments that affect the and blood vessels. Heart primary cause of fatality morbidity in people worldwide, resulting 18 million deaths per year. By identifying those who are most vulnerable to diseases ensuring they receive appropriate care, premature demise can be prevented. Machine learning algorithms now crucial medical field, especially when using databases diagnose diseases. Such efficient data processing techniques applied predict offer much potential accurate prognosis. Therefore, this study compares performance logistic regression, decision tree, support vector machine (SVM) methods with without Boruta feature selection. The Cleveland clinic dataset acquired from Kaggle, which consists 14 features 303 instances, was used investigation. It found selection algorithm, selects six relevant features, improved results algorithms. Among these classification algorithms, regression produced result, an accuracy 88.52 %.

Language: Английский

Citations

43
Diabetes risk prediction model based on community follow-up data using machine learning DOI Creative Commons
Liangjun Jiang, Zhenhua Xia, Ronghui Zhu

et al.

Preventive Medicine Reports, Journal Year: 2023, Volume and Issue: 35, P. 102358 - 102358

Published: Aug. 19, 2023

Diabetes is a chronic metabolic disease characterized by hyperglycemia, the follow-up management of diabetes patients mostly in community, but relationship between key lifestyle indicators community and risk unclear. In order to explore association life characteristic diabetes, 252,176 records people with from 2016 2023 were obtained Haizhu District, Guangzhou. According data, that affect are determined, optimal feature subset through selection technology accurately assess diabetes. A assessment model based on random forest classifier was designed, which used parameter algorithm comparison, an accuracy 91.24% AUC corresponding ROC curve 97%. improve applicability clinical real life, score card designed tested using original 95.15%, reliability high. The prediction big data mining can be for large-scale screening early warning doctors patient further promoting prevention control strategies, also wearable devices or intelligent biosensors individual self examination, reduce factor levels.

Language: Английский

Citations

25
Dynamic feature selection and quantum representation for precise heart disease prediction: Quantum-HeartDiseaseNet approach DOI

Liza M Kunjachen,

R. Kavitha

Computer Methods in Biomechanics & Biomedical Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 22

Published: Feb. 5, 2025

Cardiovascular disease is a leading cause of mortality, necessitating early and precise prediction for improved patient outcomes. This study proposes Quantum-HeartDiseaseNet, novel heart risk framework that integrates Dynamic Opposite Pufferfish Optimization Algorithm feature selection Quantum Attention-based Bidirectional Gated Recurrent Unit (QABiGRU) accurate diagnosis. The method enhances diagnosis accuracy while reducing dimensionality, Synthetic Minority Oversampling Technique (SMOTE) addresses data imbalance. Evaluated on three datasets, the proposed model achieved 98.87% accuracy, 98.74% precision, 98.56% recall, outperforming conventional methods. Experimental results validate its effectiveness in prediction.

Language: Английский

Citations

0
A Quest for Context-Specific Stock Price Prediction: A Comparison Between Time Series, Machine Learning and Deep Learning Models DOI Creative Commons
Mugdha Shailendra Kulkarni, S. Vijayakumar Bharathi, Arif Perdana

et al.

SN Computer Science, Journal Year: 2025, Volume and Issue: 6(4)

Published: April 2, 2025

Language: Английский

Citations

0
Stacking with Recursive Feature Elimination-Isolation Forest for classification of diabetes mellitus DOI Creative Commons
Nur Farahaina Idris, Mohd Arfian Ismail, M. Izham Jaya

et al.

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(5), P. e0302595 - e0302595

Published: May 8, 2024

Diabetes Mellitus is one of the oldest diseases known to humankind, dating back ancient Egypt. The disease a chronic metabolic disorder that heavily burdens healthcare providers worldwide due steady increment patients yearly. Worryingly, diabetes affects not only aging population but also children. It prevalent control this problem, as can lead many health complications. As evolution happens, humankind starts integrating computer technology with system. utilization artificial intelligence assists be more efficient in diagnosing patients, better delivery, and patient eccentric. Among advanced data mining techniques intelligence, stacking among most prominent methods applied domain. Hence, study opts investigate potential ensembles. aim reduce high complexity inherent stacking, problem contributes longer training time reduces outliers improve classification performance. In addressing concern, novel machine learning method called Stacking Recursive Feature Elimination-Isolation Forest was introduced for prediction. application Elimination design an model diagnosis while using fewer features resources. This incorporates Isolation outlier removal method. uses accuracy, precision, recall, F1 measure, time, standard deviation metrics identify performances. proposed acquired accuracy 79.077% PIMA Indians 97.446% Prediction dataset, outperforming existing demonstrating effectiveness

Language: Английский

Citations

3
Detection of Diabetes using Combined ML Algorithm DOI Open Access

Shifat Jahan Setu,

Fahima Tabassum,

M. Sarwar Jahan

et al.

International Journal of Intelligent Systems and Applications, Journal Year: 2024, Volume and Issue: 16(1), P. 11 - 23

Published: Jan. 30, 2024

Recently data clustering algorithm under machine learning are used in ‘real-life data’ to segregate them based on the outcome of a phenomenon. In this paper, diabetes is detected from pathological 768 patients using four algorithms: Fuzzy C-Means (FCM), K-means clustering, Inference system (FIS) and Support Vector Machine (SVM). Our main objective make binary classification table sense that presence or absence patient. We combined algorithms entropy-based probability enhance accuracy detection. Before applying combining scheme, we reduce size variables multiple linear regression (MLR) then logistic again applied resultant keep outlier within narrow range. Finally, entropy scheme with some modification ML got detection about 94% technique.

Language: Английский

Citations

1
Data Mining Applications in Healthcare: A Comparative Analysis of Classification Techniques for Diabetes Diagnosis Using the PIMA Indian Diabetes Dataset DOI
Prashant Verma, Aasiya Khatoon

Published: Feb. 21, 2024

Data mining involves extracting valuable information from a substantial quantity of data.Numerous applications for data can be found in different areas such as finance, healthcare & marketing. Presently, hold major significance key field precise disease prognosis and in-depth exploration medical information. Researchers employ varied minding technique help diagnose range illnesses, including cancer, diabetes, heart conditions. Diabetes has emerged modern-day with causing almost the highest number deaths around globe. It severely affects renal cardiac functions patient along eyesight loss leading to other diseases body. In this paper, we made use Kaggle's PIMA Indian Set. We deploy four strategies: logistic regression, SVM (Support Vector Machine), KNN (K-nearest neighbor), Random Forest. An assessment algorithms' performances is based on an array metrics, spanning Recall, Accuracy, Precision, F-measure. The Forest Classifier fares better compared alternatives classification methods when speaking accuracy (80.08%).

Language: Английский

Citations

1
Multi-Faceted Approach to Cardiovascular Risk Assessment by Utilizing Predictive Machine Learning and Clinical Data in a Unified Web Platform DOI Creative Commons

Khadiza Akther,

Md. Saidur Rahman Kohinoor,

Bushra Siddika Priya

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 120454 - 120473

Published: Jan. 1, 2024

Cardiovascular diseases (CVD) persist as a formidable global health challenge, underscoring the imperative for advanced early detection mechanisms. The evolution of computational methods within healthcare has paved way transformative applications machine learning, offering solutions that enhance diagnostic accuracy and contribute to SDG-3; Good Health Well-Being. This study aims identify an algorithm with consistent performance across diverse datasets integrate it into comprehensive user-centric approach heart disease prediction. investigation includes examination eight learning algorithms, three deep four heterogeneous from Kaggle. predictive these algorithms is assessed through measures include Precision, Recall, F1 score, Accuracy, Area Under Curve (AUC). A Principal Component Analysis (PCA) feature engineering presented boost performance. An alternative selection method, Lasso, was explored, PCA emerging optimal choice in given datasets. As such, XGBoost achieves impressive rate score around 99% along excellent 97% AUC prediction on other dataset. selected model integrated user-friendly web application, providing holistic platform management. Furthermore, we recommended RPA, IoTM, AI-based tailored solution make our application more reliable, which have proven attainable.

Language: Английский

Citations

1
Refining heart disease prediction accuracy using hybrid machine learning techniques with novel metaheuristic algorithms DOI
Haifeng Zhang, Rui Mu

International Journal of Cardiology, Journal Year: 2024, Volume and Issue: 416, P. 132506 - 132506

Published: Aug. 30, 2024

Language: Английский

Citations

1
A Comparative Analysis of Lung Cancer Prediction Using Machine Learning Techniques DOI

Richa Raj Srivastav,

Sachin Bhoite, Gufran Ahmad Ansari

et al.

Algorithms for intelligent systems, Journal Year: 2024, Volume and Issue: unknown, P. 457 - 476

Published: Jan. 1, 2024

Language: Английский

Citations

0