著者
足立 吉隆 田口 茂樹 弘川 奨悟
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
vol.102, no.12, pp.722-729, 2016 (Released:2016-11-30)
参考文献数
9
被引用文献数
21

Deep learning by convolution neural network (CNN) was applied to recognize a microstructure of steels. Three typical CNN-models such as LeNet5, AlexNet, and GoogLeNet were examined their accuracy of recognition. In addition to a model, an effect of learning rate, dropout ratio, and mean image subtraction on recognition accuracy were also investigated. Through this study, the potency of deep learning for microstructural classification is demonstrated.
著者
弘川 奨悟 田口 茂樹 松下 康弘 足立 吉隆
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
pp.TETSU-2017-003, (Released:2017-04-26)
参考文献数
12
被引用文献数
1

Screening important microstructure factors dominantly controlling a stress-strain curve of a dual phase steel was studied by three kinds of sparse modelling approach; Sensitive analysis, data transformation & variable selection, and Bayesian inference. In addition, an effect of data noise on descriptor screening in sparse modelling was also investigated.
著者
足立 吉隆 新川田 圭介 奥野 晃弘 弘川 奨悟 田口 茂樹 定松 直
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
pp.TETSU-2015-069, (Released:2015-10-03)
参考文献数
19
被引用文献数
2 6

Prediction of a stress-strain curve of ferrite-martensite DP steels was studied by a combined technique of Bayesian inference and artificial neural network. To screen a descriptor to be used for neural network analysis, material genomes such as volume fraction, micro-hardness, handle, and void of martensite phase, and micro-hardness of ferrite phase were examined by Bayesian inference. In a case of small data set, a machine learning method to predict mechanical properties reliably was proposed.
著者
田口 茂樹 弘川 奨悟 安田 格 徳田 耕平 足立 吉隆
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
vol.103, no.3, pp.142-148, 2017 (Released:2017-02-28)
参考文献数
8
被引用文献数
9 6

Two kinds of advanced image processing were applied to multi-phase microstructures. One is evolutional image processing where optimized filter set was suggested by genetic programing. Another is trainable WEKA segmentation where features are extracted by many kinds of filters, followed by machine learning for classification. Once an optimized filter set is determined, efficiency of image processing for new data set is improved remarkably in comparison with a case of manual image processing.
著者
足立 吉隆 新川田 圭介 奥野 晃弘 弘川 奨悟 田口 茂樹 定松 直
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
vol.102, no.1, pp.47-55, 2016 (Released:2015-12-31)
参考文献数
19
被引用文献数
6 6

Prediction of a stress-strain curve of ferrite-martensite DP steels was studied by a combined technique of Bayesian inference and artificial neural network. To screen a descriptor to be used for neural network analysis, material genomes such as volume fraction, micro-hardness, handle, and void of martensite phase, and micro-hardness of ferrite phase were examined by Bayesian inference. In a case of small data set, a machine learning method to predict mechanical properties reliably was proposed.
著者
弘川 奨悟 田口 茂樹 松下 康弘 足立 吉隆
出版者
一般社団法人 日本鉄鋼協会
雑誌
鉄と鋼 (ISSN:00211575)
巻号頁・発行日
vol.103, no.8, pp.468-474, 2017 (Released:2017-07-31)
参考文献数
12
被引用文献数
1 1

Screening important microstructure factors dominantly controlling a stress-strain curve of a dual phase steel was studied by three kinds of sparse modelling approach; Sensitive analysis, data transformation & variable selection, and Bayesian inference. In addition, an effect of data noise on descriptor screening in sparse modelling was also investigated.