Scholarlink

Skin cancer, particularly melanoma is the most virulent form of cancer, necessitating early and accurate diagnosis to enhance patient outcomes. This paper proposed an automation surgery-based ensemble approach by Convolutional Neural Network (CNN) and Inception for skin cancer detection. For the evaluation purpose ISIC 2024 data set and its accompanying metadata is used for training and testing. Following extensive hyperparameters tuning and experimentation, the CNN+ Inception model reached a maximum accuracy of 86.93% at 25 epochs. The incorporation of meta-data also enhanced predictive performance. These results highlight the promise of AI-based tools in supporting clinical decision-making. Future enhancement will be based on the optimizing the deep learning approach for better efficiency and wider clinical application.

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

Batch Size Optimization in CNN+ Inception–Based Skin Cancer Detection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Manav J. Luhar,
  • Parmanand Patel,
  • Ronak Patel,
  • Deep Kothadiya

摘要

Skin cancer, particularly melanoma is the most virulent form of cancer, necessitating early and accurate diagnosis to enhance patient outcomes. This paper proposed an automation surgery-based ensemble approach by Convolutional Neural Network (CNN) and Inception for skin cancer detection. For the evaluation purpose ISIC 2024 data set and its accompanying metadata is used for training and testing. Following extensive hyperparameters tuning and experimentation, the CNN+ Inception model reached a maximum accuracy of 86.93% at 25 epochs. The incorporation of meta-data also enhanced predictive performance. These results highlight the promise of AI-based tools in supporting clinical decision-making. Future enhancement will be based on the optimizing the deep learning approach for better efficiency and wider clinical application.