Scholarlink

<p>The given paper introduces an ensemble deep learning approach for detection of patients suffering from Glioblastoma (GBM) by combining three transfer learning models, such as MobileNetV2, DenseNet121, and ResNet50. Initially, data is pre-processed and is split into 80-10-10 format for training-validation-testing. The input images are subjected to ensemble learning models (MobileNetV2 + ResNet50 + DenseNet121). MobileNetV2 is chosen as it allows for efficient deployment on mobile and embedded devices with limited computational resources. ResNet50 is chosen as it leads to improved performance and robust features extraction while DenseNet121 prevents overfitting. The identification decision is produced by combining the outputs of three transfer models using weighted averaging done by particle swarm optimization (PSO) algorithm. The proposed ensemble framework shows the highest accuracy (95.85%), highest precision (94.37%), and highest recall (94.55%) as compared to existing studies.</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

An ensemble transfer learning metaheuristic optimization framework for detection of glioblastoma (GBM)

  • 期刊论文

  • 发表日期：

    2025年11月21日

查看全文
  • Rajasekhar Kommaraju,
  • Kurisetti Sai Sujith,
  • Yanamadala Sri Yougananda Gopi Kishore,
  • Arigala Narendra,
  • Patnala Pavan Kumar,
  • Arvapalli Home Harshavardhan,
  • G. Maheswari

摘要

The given paper introduces an ensemble deep learning approach for detection of patients suffering from Glioblastoma (GBM) by combining three transfer learning models, such as MobileNetV2, DenseNet121, and ResNet50. Initially, data is pre-processed and is split into 80-10-10 format for training-validation-testing. The input images are subjected to ensemble learning models (MobileNetV2 + ResNet50 + DenseNet121). MobileNetV2 is chosen as it allows for efficient deployment on mobile and embedded devices with limited computational resources. ResNet50 is chosen as it leads to improved performance and robust features extraction while DenseNet121 prevents overfitting. The identification decision is produced by combining the outputs of three transfer models using weighted averaging done by particle swarm optimization (PSO) algorithm. The proposed ensemble framework shows the highest accuracy (95.85%), highest precision (94.37%), and highest recall (94.55%) as compared to existing studies.