Zobrazeno 1 - 10
of 173
pro vyhledávání: '"Masanori Koyama"'
Autor:
Bunta Inoue, Masanori Koyama, Atsushi Sekiguchi, Masamitsu Shirai, Yoshihiko Hirai, Masaaki Yasuda
Publikováno v:
Journal of Photopolymer Science and Technology. 34:661-665
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::bbe22374ba4a0ff875401c180bf7b5f3
https://doi.org/10.2494/photopolymer.34.661
https://doi.org/10.2494/photopolymer.34.661
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence. 41:1979-1993
We propose a new regularization method based on virtual adversarial loss: a new measure of local smoothness of the conditional label distribution given input. Virtual adversarial loss is defined as the robustness of the conditional label distribution
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::8be1fb141fe79278d1a9dafd5df117df
https://doi.org/10.1109/tpami.2018.2858821
https://doi.org/10.1109/tpami.2018.2858821
Autor:
Yoshihiko Hirai, Masaaki Yasuda, Kosai Fukunari, Masamitsu Shirai, Hiroaki Kawata, Masanori Koyama
Publikováno v:
Journal of Photopolymer Science and Technology. 32:339-343
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::a72c44c64203a59cce98ef4090ca0fb8
https://doi.org/10.2494/photopolymer.32.339
https://doi.org/10.2494/photopolymer.32.339
Autor:
Ken Nakae, Yuji Ikegaya, Tomoe Ishikawa, Shigeyuki Oba, Hidetoshi Urakubo, Masanori Koyama, Shin Ishii
Publikováno v:
PLoS Computational Biology, Vol 10, Iss 11, p e1003949 (2014)
Crosstalk between neurons and glia may constitute a significant part of information processing in the brain. We present a novel method of statistically identifying interactions in a neuron-glia network. We attempted to identify neuron-glia interactio
Externí odkaz:
https://doaj.org/article/571fb6c4b7b44f07abf3bd9b1755ea06
Autor:
Shin-ichi Maeda, Hayato Watahiki, Yi Ouyang, Shintarou Okada, Masanori Koyama, Prabhat Nagarajan
Publikováno v:
Machine Learning and Knowledge Discovery in Databases. Research Track ISBN: 9783030865191
ECML/PKDD (2)
ECML/PKDD (2)
Practical reinforcement learning problems are often formulated as constrained Markov decision process (CMDP) problems, in which the agent has to maximize the expected return while satisfying a set of prescribed safety constraints. In this study, we c
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::85a84e8403c4f9aa0fe32ba3622bad6c
https://doi.org/10.1007/978-3-030-86520-7_35
https://doi.org/10.1007/978-3-030-86520-7_35
Publikováno v:
DLS@SC
Data parallel training is a powerful family of methods for the efficient training of deep neural networks on big data. Unfortunately, however, recent studies have shown that the merit of increased batch size in terms of both speed and model-performan
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::2098e9170cbbe68b8161f98acb038f3e
https://doi.org/10.1109/dls51937.2020.00006
https://doi.org/10.1109/dls51937.2020.00006
Autor:
Masaaki Yasuda, Yoshihiko Hirai, Masanori Koyama, Hiroaki Kawata, Reo Sakata, Masamitsu Shirai
Publikováno v:
Journal of Photopolymer Science and Technology. 31:189-192
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::4a044bb5c056fc291bf7f52a1e842c45
https://doi.org/10.2494/photopolymer.31.189
https://doi.org/10.2494/photopolymer.31.189
Publikováno v:
Japanese Journal of Applied Physics. 60:106505
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::34534afc4aa33118e0794a91170731ce
https://doi.org/10.35848/1347-4065/ac26e5
https://doi.org/10.35848/1347-4065/ac26e5
Autor:
Takeru Miyato, Masanori Koyama
Publikováno v:
Computer Vision ISBN: 9783030032432
Computer Vision
Computer Vision
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c503f8c34072ad54e38862881e7cafb8
https://doi.org/10.1007/978-3-030-03243-2_860-1
https://doi.org/10.1007/978-3-030-03243-2_860-1
Publikováno v:
KDD
The purpose of this study is to introduce new design-criteria for next-generation hyperparameter optimization software. The criteria we propose include (1) define-by-run API that allows users to construct the parameter search space dynamically, (2) e
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::25e50e44d988c0b667ca04b5ff35c59d
https://doi.org/10.1145/3292500.3330701
https://doi.org/10.1145/3292500.3330701