Scholarlink

<p>The exponential growth of data in domains such as finance, healthcare, and cybersecurity has introduced significant challenges in processing and accurately classifying large-scale, high-dimensional, and imbalanced datasets. Traditional machine learning models often struggle with scalability, adaptability, and precision when confronted with complex data structures and dynamic distributions. These limitations hinder reliable decision-making and reduce model effectiveness in real-world applications. To address these critical gaps, this study proposes an Enhanced Ensemble Learning Framework that integrates feature extraction, hyperparameter tuning, and classification into a unified approach. Specifically, a Random Forest (RF) model is dynamically fine-tuned using Particle Swarm Optimization (PSO) to adapt to varying data characteristics, improve classification accuracy, and enhance computational efficiency. The framework also incorporates a dual-layer sampling mechanism to handle class imbalance and mitigate the impact of redundant or irrelevant features. Extensive experiments conducted on benchmark datasets—including SUSY, HIGGS, MNIST, USPS, and Pendigits—demonstrate the superiority of the proposed method, achieving up to 98.3% accuracy and outperforming several state-of-the-art models in terms of robustness and scalability. These findings highlight the potential of the proposed framework as a scalable and effective solution for complex data classification tasks; however, further external validation is required to confirm its generalizability in real-world applications.</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

Optimized Ensemble Learning Framework with Enhanced Feature Selection for Efficient and Accurate Classification

  • 期刊论文

  • 发表日期：

    2025年7月22日

查看全文
  • M. L. Asha,
  • T. Johnpeter

摘要

The exponential growth of data in domains such as finance, healthcare, and cybersecurity has introduced significant challenges in processing and accurately classifying large-scale, high-dimensional, and imbalanced datasets. Traditional machine learning models often struggle with scalability, adaptability, and precision when confronted with complex data structures and dynamic distributions. These limitations hinder reliable decision-making and reduce model effectiveness in real-world applications. To address these critical gaps, this study proposes an Enhanced Ensemble Learning Framework that integrates feature extraction, hyperparameter tuning, and classification into a unified approach. Specifically, a Random Forest (RF) model is dynamically fine-tuned using Particle Swarm Optimization (PSO) to adapt to varying data characteristics, improve classification accuracy, and enhance computational efficiency. The framework also incorporates a dual-layer sampling mechanism to handle class imbalance and mitigate the impact of redundant or irrelevant features. Extensive experiments conducted on benchmark datasets—including SUSY, HIGGS, MNIST, USPS, and Pendigits—demonstrate the superiority of the proposed method, achieving up to 98.3% accuracy and outperforming several state-of-the-art models in terms of robustness and scalability. These findings highlight the potential of the proposed framework as a scalable and effective solution for complex data classification tasks; however, further external validation is required to confirm its generalizability in real-world applications.