Transfer Learning
Transfer learning is reusing knowledge a model learned on one task to improve learning on a different but related task.
What is Transfer Learning?
An image model trained on millions of general photos can be adapted to spot defects on a production line with only a few hundred examples. Transfer learning is why foundation models are so widely reusable.
How it works
A pre-trained model is used as a starting point and some or all of its layers are retrained on the new task.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions