Scholarlink

The COVID-19 pandemic, stemming from the novel coronavirus SARS-CoV-2 in late 2019, remains a persistent global public health challenge. To address this, we have developed an automated deep learning-based diagnostic model that combines features from two effective models, InceptionV3 and Xception, for COVID-19 detection. We reduce the feature set by concatenating their extracted feature vectors and applying Principal Component Analysis (PCA) for dimensionality reduction. These reduced features are input into various classifiers, including Naïve Bayes, Logistic Regression, Random Forest, KNN, and SVM. Our proposed InceptionV3 + Xception + SVM model achieves a superior classification accuracy of 99.0%, utilizing fewer features than existing models.

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 Hybrid Deep Learning Approach for COVID-19 Detection: Feature Fusion and SVM Classification

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Debendra Muduli,
  • Santosh Kumar Sharma,
  • Debasish Pradhan,
  • Madhusmita Das,
  • Suryakanta Mahapatra,
  • Saroj Kumar sahoo

摘要

The COVID-19 pandemic, stemming from the novel coronavirus SARS-CoV-2 in late 2019, remains a persistent global public health challenge. To address this, we have developed an automated deep learning-based diagnostic model that combines features from two effective models, InceptionV3 and Xception, for COVID-19 detection. We reduce the feature set by concatenating their extracted feature vectors and applying Principal Component Analysis (PCA) for dimensionality reduction. These reduced features are input into various classifiers, including Naïve Bayes, Logistic Regression, Random Forest, KNN, and SVM. Our proposed InceptionV3 + Xception + SVM model achieves a superior classification accuracy of 99.0%, utilizing fewer features than existing models.