What is transfer learning in classification models?

Transfer learning reuses pretrained models for new classification tasks.

Pretrained models like ResNet or BERT are fine-tuned on smaller datasets, saving training time and improving accuracy.