A technique in machine learning to expedite the development of new models by using a pretrained model trained on a large, relevant data set with the same inputs that can then be extended with additional layers to generate the new required outputs. The resultant new model only needs to learn the relations for your specific problem having already learnt about patterns in the data from the pretrained model. Training takes a lot of time, data, and compute resources. Sometimes a new learning task is only somewhat different from another and can benefit from transfer learning.
Glossary of Artificial Intelligence (AI), Machine Learning (ML), and Big Data Terms
Transfer Learning
