Scholarlink

The decision-making process in selecting an appropriate model for clas-classification tasks is a perplexing issue due to the variety of algorithms and datasets. This benchmarking exercise compares four popular classification algorithms concerning several benchmark datasets: Random Forests, Support Vector Classifier, Logistic Regression, and K-Nearest Neighbors. In more detail, we assess the performance of these models in multiple scenarios while checking them against conventional metrics like accuracy, precision, and recall, along with advanced ones such as hyperparameter optimization and cross-validation that use more advanced techniques. As a result of this analysis, we find that each model has both useful and counteractive features that help the real-world practitioner. Consequently, this work has relevance to the healthcare, financial, and e-commerce industries and several others by enhancing methods of class model selection, which are essential in these areas.

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

Algorithmic-Symphony: Machine Learning Classification Models Across Diverse Datasets for Performance Evaluation and Comparison

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • S. Adithya Krishna,
  • R. Sreeram,
  • Avishek Rauniyar,
  • Bowrampet Manish Reddy,
  • S. Sarath

摘要

The decision-making process in selecting an appropriate model for clas-classification tasks is a perplexing issue due to the variety of algorithms and datasets. This benchmarking exercise compares four popular classification algorithms concerning several benchmark datasets: Random Forests, Support Vector Classifier, Logistic Regression, and K-Nearest Neighbors. In more detail, we assess the performance of these models in multiple scenarios while checking them against conventional metrics like accuracy, precision, and recall, along with advanced ones such as hyperparameter optimization and cross-validation that use more advanced techniques. As a result of this analysis, we find that each model has both useful and counteractive features that help the real-world practitioner. Consequently, this work has relevance to the healthcare, financial, and e-commerce industries and several others by enhancing methods of class model selection, which are essential in these areas.