A Comprehensive Investigation of the Performances of Different Machine Learning Classifiers with SMOTE-ENN Oversampling Technique and Hyperparameter Optimization for Imbalanced Heart Failure Dataset

Abstract

Heart failure is a chronic cardiac condition characterized by reduced supply of blood to the body due to impaired contractile properties of the muscles of the heart. Like any other cardiac disorder, heart failure is a serious ailment limiting the activities and curtailing the lifespan of the patient, most often resulting in death sooner or later. Detection of survival of patients with heart failure is the path to effective intervention and good prognosis in terms of both treatment and quality of life of the patient. Machine learning techniques can be critical in this regard since they can be used to predict the survival of patients with heart failure in advance, allowing patients to receive appropriate treatment. Hence, six supervised machine learning algorithms have been studied and applied to analyze a dataset of 299 individuals from the UCI Machine Learning Repository and predict their survivability from heart failure. Three distinct approaches have been followed using Decision Tree Classifier, Logistic Regression, Gaussian Naïve Bayes, Random Forest Classifier, K-Nearest Neighbors, and Support Vector Machine algorithms. Data scaling has been performed as a preprocessing step utilizing the standard and min–max scaling method. However, grid search cross-validation and random search cross-validation techniques have been employed to optimize the hyperparameters. Additionally, the synthetic minority oversampling technique and edited nearest neighbor (SMOTE-ENN) data resampling technique are utilized, and the performances of all the approaches have been compared extensively. The experimental results clearly indicate that Random Forest Classifier (RFC) surpasses all other approaches with a test accuracy of 90% when used in combination with SMOTE-ENN and standard scaling technique. Therefore, this comprehensive investigation portrays a vivid visualization of the applicability and compatibility of different machine learning algorithms in such an imbalanced dataset and presents the role of the SMOTE-ENN algorithm and hyperparameter optimization for enhancing the performances of the machine learning algorithms.

Evaluation of feature scaling for improving the performance of supervised learning methods

“…This study employed an HF dataset obtained from the Institute of Cardiology and Allied hospital by Ahmad et al [21] previously studied by [22]- [25] for HF survival prediction. The dataset contains 299 samples of patients aged above 40 years.…”

Efficiency due to data balancing in the prediction of essential proteins

“…The above-mentioned method was proposed by [33] to predict the survivability of heart failure where the dataset was collected from the UCI dataset [4] and applied to six different classifiers such as Decision Tree Classifier, Logistic Regression, Gaussian Naïve Bayes, Random Forest classifier, K-nearest Neighbors (KNN), and Support Vector Machine (SVM). As the higher weight values dominate the machine learning classifier than the lower weight values, the proposed method scaled the dataset using the min-max scaling method.…”

“…Normalization means standardizing the individual features of the dataset. Normalization helps to make the features bias less as the machine learning algorithm gives higher or lower weight to the features at the time of learning [33]. With the help of normalization, feature values get the mean of zero and the variance of one.…”

“…Therefore, to improve the precision of the minority class, data balancing a very significant task. Only a few of the algorithms can accurately inherit the characteristics because of the lack of information on the minor class and as a result, the classification model's decision boundary is greatly compressed [33]. In statistical learning, the learning shows the worse result for imbalanced data and in the real-world application, it is observed that one class is larger than the other class [24].…”

Rough-Fuzzy Based Synthetic Data Generation Exploring Boundary Region of Rough Sets to Handle Class Imbalance Problem

“…For accurate prediction, the researchers developed a similarity classifier in their work [31] to address the Archimedean-Dombi aggregation operators, which are well known for offering sufficient generalization in aggregating data. There are many crisp oversampling techniques in the literature [32][33][34][35][36][37], which may not handle the class imbalance problem properly due to the presence of some uncertainties in the data. Fuzzy theory is very popularly used for solving such problems.…”