paper-with-me

Papers

Maximum Mean Discrepancy for Generalization in the Presence of Distribution and Missingness Shift

2021-11-19 · Liwen Ouyang, Aaron Key

Covariate shifts are a common problem in predictive modeling on real-world problems. This paper proposes addressing the covariate shift problem by minimizing Maximum Mean Discrepancy (MMD) statistics between the training and test sets in either feature input space, feature representation space, or both. We designed three techniques that we call MMD Representation, MMD Mask, and MMD Hybrid to deal with the scenarios where only a distribution shift exists, only a missingness shift exists, or both types of shift exist, respectively. We find that integrating an MMD loss component helps models use the best features for generalization and avoid dangerous extrapolation as much as possible for each test sample. Models treated with this MMD approach show better performance, calibration, and extrapolation on the test set.

📄 PDF Abstract BibTeX arXiv:2111.10344

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Stein Discrepancy for Unsupervised Domain Adaptation

2025-02-05 · Anneke von Seeger, Dongmian Zou, Gilad Lerman

Unsupervised domain adaptation (UDA) leverages information from a labeled source dataset to improve accuracy on a related but unlabeled target dataset. A common approach to UDA is aligning representations from the source…

Domain AdaptationUnsupervised Domain Adaptation

Permutation Complexity Bound on Out-Sample Error

2010-12-01 · NeurIPS 2010 12 · Malik Magdon-Ismail

We define a data dependent permutation complexity for a hypothesis set \math{\hset}, which is similar to a Rademacher complexity or maximum discrepancy. The permutation complexity is based like the maximum discrepancy on…

Math

Training generative neural networks via Maximum Mean Discrepancy optimization

2015-05-14 · Gintare Karolina Dziugaite, Daniel M. Roy, Zoubin Ghahramani

We consider training a deep neural network to generate samples from an unknown distribution given i.i.d. data. We frame learning as an optimization minimizing a two-sample test statistic---informally speaking, a good gen…

On Minimum Discrepancy Estimation for Deep Domain Adaptation

2019-01-02 · Mohammad Mahfujur Rahman, Clinton Fookes, Mahsa Baktashmotlagh, Sridha Sridharan

In the presence of large sets of labeled data, Deep Learning (DL) has accomplished extraordinary triumphs in the avenue of computer vision, particularly in object classification and recognition tasks. However, DL cannot …

Domain AdaptationGeneral Classificationimage-classificationImage Classification

Learning Kernel Tests Without Data Splitting

2020-06-03 · NeurIPS 2020 12 · Jonas M. Kübler, Wittawat Jitkrittum, Bernhard Schölkopf, Krikamol Muandet

Modern large-scale kernel-based tests such as maximum mean discrepancy (MMD) and kernelized Stein discrepancy (KSD) optimize kernel hyperparameters on a held-out sample via data splitting to obtain the most powerful test…