Scholarlink

Transfer learning (TL) is a powerful technique in machine learning. Here, a pre-trained model carrying experience with similar types of tasks is applied to improve the efficacy of the base model with similar tasks. In current days, TL is widely applied in disease classification, identification of malignant cells and natural language processing. It is a promising area with high potential. The domain shifts, negative transfer, and overfitting are the major challenges and limitations of TL. In future, TL will overcome its limitations and will be widely applied in pathology. The present chapter discusses the basic principles, strengths, limitations and applications of transfer learning.

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

Transfer Learning: Leveraging Pre-trained Knowledge for Efficient Models in Deep Learning

  • 章节

  • 发表日期：

    2025年

查看全文
  • Pranab Dey

摘要

Transfer learning (TL) is a powerful technique in machine learning. Here, a pre-trained model carrying experience with similar types of tasks is applied to improve the efficacy of the base model with similar tasks. In current days, TL is widely applied in disease classification, identification of malignant cells and natural language processing. It is a promising area with high potential. The domain shifts, negative transfer, and overfitting are the major challenges and limitations of TL. In future, TL will overcome its limitations and will be widely applied in pathology. The present chapter discusses the basic principles, strengths, limitations and applications of transfer learning.