Scholarlink

<p>The Synthetic Minority Oversampling Technique (SMOTE) has emerged as an effective and commonly used method for addressing class imbalance in machine learning modeling. This survey investigates the statistical variations of SMOTE, highlighting its enhancement types and techniques. Examining the statistical variants of SMOTE, specifically how synthetic instance production affects model performance, bias, overfitting, and generalization are the main ones. The research also emphasizes recent improvements, such as selection- and adjustment-based approaches. Overfitting, high-dimensional data, and generalization concerns are all covered. This survey provides an in-depth review of the usefulness and limitations of statistically based SMOTE and advises researchers and practitioners on selecting and customizing SMOTE techniques for specific datasets.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A comprehensive review of statistical variants and enhancements of SMOTE oversampling method

  • 期刊论文

  • 发表日期：

    2025年8月1日

查看全文
  • Sahar Idwan,
  • Wael Etaiwi,
  • Haneen Rafayia,
  • Izzeddin Matar

摘要

The Synthetic Minority Oversampling Technique (SMOTE) has emerged as an effective and commonly used method for addressing class imbalance in machine learning modeling. This survey investigates the statistical variations of SMOTE, highlighting its enhancement types and techniques. Examining the statistical variants of SMOTE, specifically how synthetic instance production affects model performance, bias, overfitting, and generalization are the main ones. The research also emphasizes recent improvements, such as selection- and adjustment-based approaches. Overfitting, high-dimensional data, and generalization concerns are all covered. This survey provides an in-depth review of the usefulness and limitations of statistically based SMOTE and advises researchers and practitioners on selecting and customizing SMOTE techniques for specific datasets.