Chances are you are familiar with the recent and seemingly endless machine learning innovations, but do you know about what goes into training a machine learning model? Generally, a given machine learning model is trained on specific data for a particular task. This training process can be exceptionally resource and time-consuming, and since the resulting models are task-specific, the maximum potential of the resulting model is not realized.

Optimally-performing neural network models, for example, are often the result of many iterations of fine-tuning from researchers or practitioners. Could these trained models not be additionally exploited for a wider assortment of tasks? Transfer learning involves the leveraging of existing machine learning models for use in scenarios in which the models were not originally trained.

Much as humans do not discard everything they have previously learned and start a fresh each time they take up a new task, transfer learning allows a machine learning model to port the knowledge it has acquired during training to new tasks, extending the reach of the combination of computation and expertise having been used as fuel for the original model. Simply put, transfer learning can save training time and extend the usefulness of existing machine learning models. It is also an invaluable technique for tasks where the large amounts of training data typically required for training a model from scratch are not available.

Becoming familiar with complex concepts and implementing these concepts in practice are two very different things, and this is where Hands-On Transfer Learning with Python shines. The book starts with a deep dive into both deep learning and transfer learning, conceptually. This is followed by practical implementations of these concepts with real-world examples and research problems, using modern deep learning tools from the Python ecosystem, such as TensorFlow and Keras. Dipanjan, Raghav, and Tamoghna excel at elegantly marrying the theoretical and the practical, a remarkable advantage for the reader of such a well-crafted publication.

Transfer learning has shown much promise of late in many domains, and is a very active area of contemporary machine learning research. If you are looking for a complete guide to both deep learning and transfer learning, starting from zero, Hands-On Transfer Learning with Python should be your first stop.

Matthew Mayo

Editor, KDnuggets

@mattmayo13