Recycling AI Algorithms with Transfer Learning

Algorithm developers are using transfer learning to reuse the experience gained by one algorithm as the starting point for building another one for performing related tasks. Humans can transfer their knowledge across different tasks. For instance, if a person knows how to ride a bike, then he can easily transfer his knowledge to learn how to drive a car. Transfer learning is a similar concept. It is a process that enables developers to use the experience gained by one model while performing one task and apply it to a second model to solve different but related tasks.