AI Glossary/Transfer Learning
Deep Learning

Transfer Learning

Using knowledge learned from one task to improve performance on a different but related task.

In-depth explanation

Transfer learning leverages pre-trained models, typically trained on large datasets, as starting points for new tasks. Fine-tuning adapts the pre-trained weights to the target task. This approach dramatically reduces data and compute requirements for new tasks. It's especially powerful in computer vision and NLP where pre-training is expensive.

Examples

Using ImageNet-pretrained models
Fine-tuning BERT for sentiment

Related terms

Master Transfer Learning.

Learn how to apply this concept with hands-on projects in our comprehensive AI programs.