Estimating the Transfer Learning Ability of a Deep Neural Networks by Means of Representations

Magai German, Soroka A.

The basis of transfer learning methods is the ability of deep neural networks to use knowledge from one domain to learn in another domain. However, another important task is the analysis and explanation of the internal representations of deep neural networks models in the process of transfer learning. Some deep models are known to be better at transferring knowledge than others. In this research, we apply the Centered Kernel Alignment (CKA) method to analyze the internal representations of deep neural networks and propose a method to evaluate the ability of a neural network architecture to transfer knowledge based on the quantitative change in representations during the learning process. We introduce the Transfer Ability Score (TAs) measure to assess the ability of an architecture to effectively transfer learning. We test our approach using Vision Transformer (ViT-B/16) and CNN (ResNet, DenseNet) architectures in computer vision tasks in several datasets, including medical images. Our work is an attempt to explain the transfer learning process.

Keywords: Transfer Learning, knowledge representation and reasoning, explainable AI, Deep learning architectures and techniques

In book

Advances in Neural Computation, Machine Learning, and Cognitive Research VII

Magaj G., Soroka A. Vol. 1120. , Studies in Computational Intelligence, 2023.