欠采样
随机森林
决策树
阿达布思
计算机科学
朴素贝叶斯分类器
特征选择
人工智能
过采样
机器学习
逻辑回归
领域(数学)
数据挖掘
单变量
模式识别(心理学)
支持向量机
数学
多元统计
计算机网络
纯数学
带宽(计算)
作者
Saravanan Parthasarathy,Vaishnavi Jayaraman,Jane Preetha Princy R
标识
DOI:10.1109/idciot56793.2023.10053458
摘要
cardiovascular diseases rank among the top causes of death around the world. Anticipating cardiovascular illness is a major challenge for the healthcare industry. It has been demonstrated that the implementation of Machine Learning (ML), Artificial Intelligence (AI), and data science may effectively aid in decision-making and prediction using the huge quantities of data created by the healthcare industry. The medical field has profited immensely from the use of algorithms and correlation approaches for identifying patterns in the vitals. An imbalanced heart failure data set was analyzed using Logistic Regression, Naive Bayes, Decision Tree, AdaBoost, Random Forest, and XGBoost (XGB). The univariate feature selection model f_classif was used to identify the most relevant characteristics after the dataset was normalized using the Z-score method. This dataset was then balanced by oversampling and undersampling with SMOTE-ENN. Compared to the other ML models applied to the balanced dataset, XGBoost achieved higher levels of accuracy (97%), precision (96%), recall (96%), and F1-score (96%) in classifying heart failure.
科研通智能强力驱动
Strongly Powered by AbleSci AI