paper-with-me

홈 › Papers

A Riemannian Mean Field Formulation for Two-layer Neural Networks with Batch Normalization

2021-10-17 · Chao Ma, Lexing Ying

The training dynamics of two-layer neural networks with batch normalization (BN) is studied. It is written as the training dynamics of a neural network without BN on a Riemannian manifold. Therefore, we identify BN's effect of changing the metric in the parameter space. Later, the infinite-width limit of the two-layer neural networks with BN is considered, and a mean-field formulation is derived for the training dynamics. The training dynamics of the mean-field formulation is shown to be the Wasserstein gradient flow on the manifold. Theoretical analysis are provided on the well-posedness and convergence of the Wasserstein gradient flow.

📄 PDF Abstract BibTeX arXiv:2110.08725

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Batch Normalization 설명 없음

Similar Papers 제목 키워드 기반

Riemannian Batch Normalization: A Gyro Approach

2025-09-08 · Ziheng Chen, Xiao-Jun Wu, Bernhard Schölkopf, Nicu Sebe arxiv

Normalization layers are crucial for deep learning, but their Euclidean formulations are inadequate for data on manifolds. On the other hand, many Riemannian manifolds in machine learning admit gyro-structures, enabling …

An Alternative to EM for Gaussian Mixture Models: Batch and Stochastic Riemannian Optimization

2017-06-10 · Reshad Hosseini, Suvrit Sra

We consider maximum likelihood estimation for Gaussian Mixture Models (Gmms). This task is almost invariably solved (in theory and practice) via the Expectation Maximization (EM) algorithm. EM owes its success to various…

Density Estimationparameter estimationRiemannian optimization

Batch Normalization for Neural Networks on Complex Domains

2026-05-01 · Xuan Son Nguyen, Nistor Grozavu arxiv

Riemannian neural networks have proven effective in solving a variety of machine learning tasks. The key to their success lies in the development of principled Riemannian analogs of fundamental building blocks in deep ne…

Node ClassificationAction Recognition

Riemannian batch normalization for SPD neural networks

2019-09-03 · NeurIPS 2019 12 · Daniel Brooks, Olivier Schwander, Frederic Barbaresco, Jean-Yves Schneider 외

Covariance matrices have attracted attention for machine learning applications due to their capacity to capture interesting structure in the data. The main challenge is that one needs to take into account the particular …

Action Recognition

A Lie Group Approach to Riemannian Batch Normalization

2024-03-17 · Ziheng Chen, Yue Song, Yunmei Liu, Nicu Sebe

Manifold-valued measurements exist in numerous applications within computer vision and machine learning. Recent studies have extended Deep Neural Networks (DNNs) to manifolds, and concomitantly, normalization techniques …

Action RecognitionEEGTemporal Action Localization