Refining generative process with discriminator guidance in score-based diffusion models

D Kim, Y Kim, SJ Kwon, W Kang, IC Moon - arxiv preprint arxiv …, 2022‏ - arxiv.org
The proposed method, Discriminator Guidance, aims to improve sample generation of pre-
trained diffusion models. The approach introduces a discriminator that gives explicit …

Adapting to continuous covariate shift via online density ratio estimation

YJ Zhang, ZY Zhang, P Zhao… - Advances in Neural …, 2023‏ - proceedings.neurips.cc
Dealing with distribution shifts is one of the central challenges for modern machine learning.
One fundamental situation is the covariate shift, where the input distributions of data change …

Towards a unified analysis of kernel-based methods under covariate shift

X Feng, X He, C Wang, C Wang… - Advances in Neural …, 2023‏ - proceedings.neurips.cc
Covariate shift occurs prevalently in practice, where the input distributions of the source and
target data are substantially different. Despite its practical importance in various learning …

Estimating the density ratio between distributions with high discrepancy using multinomial logistic regression

A Srivastava, S Han, K Xu, B Rhodes… - arxiv preprint arxiv …, 2023‏ - arxiv.org
Functions of the ratio of the densities $ p/q $ are widely used in machine learning to quantify
the discrepancy between the two distributions $ p $ and $ q $. For high-dimensional …

Overcoming saturation in density ratio estimation by iterated regularization

L Gruber, M Holzleitner, J Lehner, S Hochreiter… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Estimating the ratio of two probability densities from finitely many samples, is a central task
in machine learning and statistics. In this work, we show that a large class of kernel methods …

Meta-learning for relative density-ratio estimation

A Kumagai, T Iwata, Y Fujiwara - Advances in Neural …, 2021‏ - proceedings.neurips.cc
The ratio of two probability densities, called a density-ratio, is a vital quantity in machine
learning. In particular, a relative density-ratio, which is a bounded extension of the density …

Revealing Distribution Discrepancy by Sampling Transfer in Unlabeled Data

Z Zhao, L Cao, X Fan… - Advances in Neural …, 2025‏ - proceedings.neurips.cc
There are increasing cases where the class labels of test samples are unavailable, creating
a significant need and challenge in measuring the discrepancy between training and test …

Computing high-dimensional optimal transport by flow neural networks

C Xu, X Cheng, Y **e - arxiv preprint arxiv:2305.11857, 2023‏ - arxiv.org
Flow-based models are widely used in generative tasks, including normalizing flow, where a
neural network transports from a data distribution $ P $ to a normal distribution. This work …

Deep bregman divergence for self-supervised representations learning

M Rezaei, F Soleymani, B Bischl, S Azizi - Computer Vision and Image …, 2023‏ - Elsevier
Neural Bregman divergence measures the divergence of data points using convex neural
networks, which is beyond Euclidean distance and capable of capturing divergence over …

Federated learning under covariate shifts with generalization guarantees

A Ramezani-Kebrya, F Liu, T Pethick… - arxiv preprint arxiv …, 2023‏ - arxiv.org
This paper addresses intra-client and inter-client covariate shifts in federated learning (FL)
with a focus on the overall generalization performance. To handle covariate shifts, we …