Transfer Learning

Friendly Description: Transfer learning is when an AI uses what it learned doing one task to get a head start on another. It's like a great pianist learning the organ: most of the skills carry over, so they pick it up much faster than someone starting from scratch. Transfer learning saves enormous amounts of time and data when building new AI applications.

Example: A model trained on millions of general photos can be transferred to medical imaging by giving it a much smaller set of X-ray examples. It already knows how to interpret edges, shapes, and shading, so it learns the medical task far quicker than a model starting from zero.