用极端梯度增强来对胰腺癌进行分类的堆叠集团深度学习

Wilson Bakasa1, Serestina Viriri1

  • 1School of Mathematics Statistics & Computer Science, College of Agriculture, Engineering and Science, University of KwaZulu-Natal, Durban, South Africa.

概括

堆叠组合深度学习 (SEDL) 通过结合多个模型来增强胰腺CT图像分类. 这种组合方法实现了98.8%的准确性,改善了诊断管道的性能.