关注
Ryo Karakida
Ryo Karakida
AIST (National Institute of Advanced Industrial Science and Technology)
在 aist.go.jp 的电子邮件经过验证 - 首页
标题
引用次数
引用次数
年份
Universal statistics of Fisher information in deep neural networks: mean field approach
R Karakida, S Akaho, S Amari
International Conference on Artificial Intelligence and Statistics (AISTATS …, 2018
1382018
Information geometry connecting Wasserstein distance and Kullback–Leibler divergence via the entropy-relaxed transportation problem
S Amari, R Karakida, M Oizumi
Information Geometry 1, 13-37, 2018
772018
Dynamical analysis of contrastive divergence learning: Restricted Boltzmann machines with Gaussian visible units
R Karakida, M Okada, S Amari
Neural Networks 79, 78-87, 2016
552016
Fisher information and natural gradient learning in random deep networks
S Amari, R Karakida, M Oizumi
International Conference on Artificial Intelligence and Statistics, 694-702, 2019
422019
The normalization method for alleviating pathological sharpness in wide neural networks
R Karakida, S Akaho, S Amari
Advances in Neural Information Processing Systems, 6406--6416, 2019
412019
Dynamics of learning in MLP: Natural gradient and singularity revisited
S Amari, T Ozeki, R Karakida, Y Yoshida, M Okada
Neural computation 30 (1), 1-33, 2017
352017
Pathological spectra of the Fisher information metric and its variants in deep neural networks
R Karakida, S Akaho, S Amari
Neural Computation 33 (8), 2274-2307, 2021
332021
Understanding Approximate Fisher Information for Fast Convergence of Natural Gradient Descent in Wide Neural Networks
R Karakida, K Osawa
Advances in Neural Information Processing Systems (NeurIPS), 2020
262020
Information geometry for regularized optimal transport and barycenters of patterns
S Amari, R Karakida, M Oizumi, M Cuturi
Neural computation 31 (5), 827-848, 2019
212019
Statistical mechanical analysis of online learning with weight normalization in single layer perceptron
Y Yoshida, R Karakida, M Okada, S Amari
Journal of the Physical Society of Japan 86 (4), 044002, 2017
172017
Self-paced data augmentation for training neural networks
T Takase, R Karakida, H Asoh
Neurocomputing 442, 296-306, 2021
162021
Learning curves for continual learning in neural networks: Self-knowledge transfer and forgetting
R Karakida, S Akaho
International Conference on Learning Representations, 2022
122022
Statistical mechanical analysis of learning dynamics of two-layer perceptron with multiple output units
Y Yoshida, R Karakida, M Okada, SI Amari
Journal of Physics A: Mathematical and Theoretical 52 (18), 184002, 2019
112019
Statistical neurodynamics of deep networks: Geometry of signal spaces
S Amari, R Karakida, M Oizumi
Nonlinear Theory and Its Applications, IEICE 10 (4), 322-336, 2019
102019
Understanding gradient regularization in deep learning: Efficient finite-difference computation and implicit bias
R Karakida, T Takase, T Hayase, K Osawa
International Conference on Machine Learning, 15809-15827, 2023
82023
The spectrum of Fisher information of deep networks achieving dynamical isometry
T Hayase, R Karakida
International Conference on Artificial Intelligence and Statistics, 334-342, 2021
72021
Adaptive Natural Gradient Learning Algorithms for Unnormalized Statistical Models
R Karakida, M Okada, S Amari
Proceedings of International Conference on Artificial Neural Networks (ICANN …, 2016
72016
Analyzing feature extraction by contrastive divergence learning in RBMs
R Karakida, M Okada, S Amari
Deep learning and representation learning workshop: NIPS, 2014
72014
Information geometry of wasserstein divergence
R Karakida, S Amari
International Conference on Geometric Science of Information, 119-126, 2017
52017
Maximum likelihood learning of RBMs with Gaussian visible units on the Stiefel manifold
R Karakida, M Okada, S Amari
Proceedings of 24th European Symposium on Artificial Neural Networks …, 2016
32016
系统目前无法执行此操作,请稍后再试。
文章 1–20