AM 207 Final Project
Group: Dashiell Young-Saver, Morris Reeves, Manana Hakobyan, Blake Bullwinkel
This repository contains a TensorFlow implementation of the model for transfer learning proposed in Fast Adaptation with Linearized Neural Networks by Maddox et al. (2021), along with experiments using synthetic and real-world data that highlight scenarios in which the proposed method fails to outperform the predictive performance of GPs with generic RBF kernels.
Original paper: http://proceedings.mlr.press/v130/maddox21a/maddox21a.pdf