Dynamic Flows on Curved Space Generated by Labeled Data
Refereed conference paper presented and published in conference proceedings

Full Text

Other information
AbstractThe scarcity of labeled data is a long-standing challenge for many machine learning tasks. We propose our gradient flow method to leverage the existing dataset (i.e., source) to generate new samples that are close to the dataset of interest (i.e., target). We lift both datasets to the space of probability distributions on the feature-Gaussian manifold, and then develop a gradient flow method that minimizes the maximum mean discrepancy loss. To perform the gradient flow of distributions on the curved feature-Gaussian space, we unravel the Riemannian structure of the space and compute explicitly the Riemannian gradient of the loss function induced by the optimal transport metric. For practical applications, we also propose a discretized flow, and provide conditional results guaranteeing the global convergence of the flow to the optimum. We illustrate the results of our proposed gradient flow method on several real-world datasets and show our method can improve the accuracy of classification models in transfer learning settings.
All Author(s) ListXinru Hua, Truyen Nguyen, Tam Le, Jose Blanchet, Viet Anh Nguyen
Name of Conference32nd International Joint Conference on Artificial Intelligence
Start Date of Conference19/08/2023
End Date of Conference25/08/2023
Place of ConferenceMacao
Country/Region of ConferenceMacau
LanguagesEnglish-United States

Last updated on 2023-14-09 at 12:29