Chako Takahashi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Chako Takahashi has a total of 81 co-authors. Here are the top co-authors by number of shared publications:
- Muneki Yasuda
Graduate School of Science and Engineering, Yamagata University
5 shared publications - Katsumi Ida
National Institute for Fusion Science, National Institutes of Natural Sciences 1 , Toki 509-5292
4 shared publications - Keisuke Matsuoka
National Institute for Fusion Science, Toki 509-5292, Japan
4 shared publications - Shoichi Okamura
National Institute for Fusion Science, National Institutes of Natural Sciences
4 shared publications H. Iguchi
National Institute for Fusion Science, Toki 509-5292, Japan
3 shared publications- Mitsutaka Isobe
National Institutes of Natural Sciences
3 shared publications
Latest Publications
- 26 February 2025
- Published by Elsevier in Neural Networks
- Vol. 187, 107297
Dataset-free weight-initialization on restricted Boltzmann machine
- 26 February 2025
- Published by Elsevier in Neural Networks
- Vol. 187, 107297
Dataset-free weight-initialization on restricted Boltzmann machine
- Open Access
- 1 January 2025
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 16 (3) , 410-421
Exact solvable deep Boltzmann machine
In training and inference in a deep Boltzmann machine (DBM), the evaluation of the free energy is crucial. Due to combinatorial explosion, its evaluation is computationally infeasible. We investigate the free energy evaluation of the DBM based on the Bethe free energy representation, which is related to the belief propagation algorithm. This representation significantly reduces the computational cost of the evaluation and enables the exact computation of the free energy in a slim DBM, i.e., a DBM with small-sized layers. Finally, we discuss the properties of the free energy in slim DBMs based on numerical and analytical methods.
- 1 January 2025
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 16 (3) , 410-421
Open AccessExact solvable deep Boltzmann machine
In training and inference in a deep Boltzmann machine (DBM), the evaluation of the free energy is crucial. Due to combinatorial explosion, its evaluation is computationally infeasible. We investigate the free energy evaluation of the DBM based on the Bethe free energy representation, which is related to the belief propagation algorithm. This representation significantly reduces the computational cost of the evaluation and enables the exact computation of the free energy in a slim DBM, i.e., a DBM with small-sized layers. Finally, we discuss the properties of the free energy in slim DBMs based on numerical and analytical methods.
- Open Access
- 2 September 2024
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 63 (9) , 09SP03
Deep learning model for predicting the spatial distribution of binding energy from atomic configurations
Understanding plasma-material interaction is crucial for achieving steady-state operation of magnetic confinement fusion devices. Kinetic Monte Carlo (kMC) simulation is a powerful tool for investigating the motion of atoms in the plasma facing materials under the influence of this interaction. To predict trapping sites and migration energies necessary for kMC simulations, we developed a deep learning model based on pix2pix for predicting the spatial distribution of binding energy. Results show that the model can reproduce spatial distributions similar to the true values. However, larger errors occur in regions with steep value gradients.
- 2 September 2024
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 63 (9) , 09SP03
Open AccessDeep learning model for predicting the spatial distribution of binding energy from atomic configurations
Understanding plasma-material interaction is crucial for achieving steady-state operation of magnetic confinement fusion devices. Kinetic Monte Carlo (kMC) simulation is a powerful tool for investigating the motion of atoms in the plasma facing materials under the influence of this interaction. To predict trapping sites and migration energies necessary for kMC simulations, we developed a deep learning model based on pix2pix for predicting the spatial distribution of binding energy. Results show that the model can reproduce spatial distributions similar to the true values. However, larger errors occur in regions with steep value gradients.
- 23 August 2024
- Published by Springer Nature
- p. 244-255
Pool-Based Active Classification Based on Expected Error Reduction with Uncertainty Subsampling
- 23 August 2024
- Published by Springer Nature
- p. 244-255
Pool-Based Active Classification Based on Expected Error Reduction with Uncertainty Subsampling
- Open Access
- 1 January 2024
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 15 (2) , 273-283
Quasi-free energy evaluation of Gaussian-Bernoulli restricted Boltzmann machine for anomaly detection
A Gaussian-Bernoulli restricted Boltzmann machine (GBRBM) is often used in semi-supervised anomaly detection (AD), in which the GBRBM is trained using only normal data points. The GBRBM-based AD is performed based on a score that is identical with an energy function of the marginalized GBRBM. However, it is difficult to set a threshold of the score for discriminating between a normal and an anomaly to an appropriate value because we do not equip a valid interpretation for the score value. To gain the interpretation, we focus on features of the score: the average, variance, and minimum values; and propose a sampling-based method for evaluating the features. Numerical experiments demonstrate that the proposed method can evaluate these three quantities with a high accuracy.
- 1 January 2024
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 15 (2) , 273-283
Open AccessQuasi-free energy evaluation of Gaussian-Bernoulli restricted Boltzmann machine for anomaly detection
A Gaussian-Bernoulli restricted Boltzmann machine (GBRBM) is often used in semi-supervised anomaly detection (AD), in which the GBRBM is trained using only normal data points. The GBRBM-based AD is performed based on a score that is identical with an energy function of the marginalized GBRBM. However, it is difficult to set a threshold of the score for discriminating between a normal and an anomaly to an appropriate value because we do not equip a valid interpretation for the score value. To gain the interpretation, we focus on features of the score: the average, variance, and minimum values; and propose a sampling-based method for evaluating the features. Numerical experiments demonstrate that the proposed method can evaluate these three quantities with a high accuracy.
- Open Access
- 1 January 2023
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 14 (2) , 491-499
Backdoor poisoning attacks against few-shot classifiers based on meta-learning
Few-shot classification is a classification made on the basis of very few samples, and meta-learning methods (also called “learning to learn”) are often employed to accomplish it. Research on poisoning attacks against meta-learning-based few-shot classifier has recently started to be investigated. While poisoning attacks aimed at disrupting the availability of the classifier during meta-testing have been studied in Xu et al. [1] and Oldewage et al. [2], backdoor poisoning in meta-testing has only been briefly explored by Oldewage et al. [2] under limited conditions. We formulate a backdoor poisoning attack on meta-learning-based few-shot classification in this study. We show that the proposed backdoor poisoning attack is effective against the few-shot classification using model-agnostic meta-learning (MAML) [3] through experiments.
- 1 January 2023
- Published by Institute of Electronics, Information and Communications Engineers (IEICE) in Nonlinear Theory and Its Applications, IEICE
- Vol. 14 (2) , 491-499
Open AccessBackdoor poisoning attacks against few-shot classifiers based on meta-learning
Few-shot classification is a classification made on the basis of very few samples, and meta-learning methods (also called “learning to learn”) are often employed to accomplish it. Research on poisoning attacks against meta-learning-based few-shot classifier has recently started to be investigated. While poisoning attacks aimed at disrupting the availability of the classifier during meta-testing have been studied in Xu et al. [1] and Oldewage et al. [2], backdoor poisoning in meta-testing has only been briefly explored by Oldewage et al. [2] under limited conditions. We formulate a backdoor poisoning attack on meta-learning-based few-shot classification in this study. We show that the proposed backdoor poisoning attack is effective against the few-shot classification using model-agnostic meta-learning (MAML) [3] through experiments.
- 25 August 2022
- Published by American Physical Society (APS) in Physical Review E
- Vol. 106 (2) , 024127
Free energy evaluation using marginalized annealed importance sampling
The evaluation of the free energy of a stochastic model is considered a significant issue in various fields of physics and machine learning. However, the exact free energy evaluation is computationally infeasible because the free energy expression includes an intractable partition function. Annealed importance sampling (AIS) is a type of importance sampling based on the Markov chain Monte Carlo method that is similar to a simulated annealing and can effectively approximate the free energy. This study proposes an AIS-based approach, which is referred to as marginalized AIS (mAIS). The statistical efficiency of mAIS is investigated in detail based on theoretical and numerical perspectives. Based on the investigation, it is proved that mAIS is more effective than AIS under a certain condition.
- 25 August 2022
- Published by American Physical Society (APS) in Physical Review E
- Vol. 106 (2) , 024127
Free energy evaluation using marginalized annealed importance sampling
The evaluation of the free energy of a stochastic model is considered a significant issue in various fields of physics and machine learning. However, the exact free energy evaluation is computationally infeasible because the free energy expression includes an intractable partition function. Annealed importance sampling (AIS) is a type of importance sampling based on the Markov chain Monte Carlo method that is similar to a simulated annealing and can effectively approximate the free energy. This study proposes an AIS-based approach, which is referred to as marginalized AIS (mAIS). The statistical efficiency of mAIS is investigated in detail based on theoretical and numerical perspectives. Based on the investigation, it is proved that mAIS is more effective than AIS under a certain condition.
- Open Access
Free Energy Evaluation Using Marginalized Annealed Importance Sampling
The evaluation of the free energy of a stochastic model is considered a significant issue in various fields of physics and machine learning. However, the exact free energy evaluation is computationally infeasible because the free energy expression includes an intractable partition function. Annealed importance sampling (AIS) is a type of importance sampling based on the Markov chain Monte Carlo method that is similar to a simulated annealing and can effectively approximate the free energy. This study proposes an AIS-based approach, which is referred to as marginalized AIS (mAIS). The statistical efficiency of mAIS is investigated in detail based on theoretical and numerical perspectives. Based on the investigation, it is proved that mAIS is more effective than AIS under a certain condition.
Open AccessFree Energy Evaluation Using Marginalized Annealed Importance Sampling
The evaluation of the free energy of a stochastic model is considered a significant issue in various fields of physics and machine learning. However, the exact free energy evaluation is computationally infeasible because the free energy expression includes an intractable partition function. Annealed importance sampling (AIS) is a type of importance sampling based on the Markov chain Monte Carlo method that is similar to a simulated annealing and can effectively approximate the free energy. This study proposes an AIS-based approach, which is referred to as marginalized AIS (mAIS). The statistical efficiency of mAIS is investigated in detail based on theoretical and numerical perspectives. Based on the investigation, it is proved that mAIS is more effective than AIS under a certain condition.
- 29 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Targeted Universal Adversarial Examples via Single Node Injection on Graph Convolutional Networks
Graph convolutional networks (GCNs) are fundamental graph neural networks used for solving node classification problems in graph-structured data. GCNs have been reported to be vulnerable to adversarial example attacks, posing a severe threat to their practical applications. In this study, we formulate targeted universal adversarial example attacks performed by injecting a single fake node into a graph. The proposed methods eliminate an unrealistic assumption required for previously proposed universal adversarial example attacks on GCNs. We show that GCNs are highly vulnerable to universal adversarial examples generated by only injecting a single node through experiments.
- 29 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Targeted Universal Adversarial Examples via Single Node Injection on Graph Convolutional Networks
Graph convolutional networks (GCNs) are fundamental graph neural networks used for solving node classification problems in graph-structured data. GCNs have been reported to be vulnerable to adversarial example attacks, posing a severe threat to their practical applications. In this study, we formulate targeted universal adversarial example attacks performed by injecting a single fake node into a graph. The proposed methods eliminate an unrealistic assumption required for previously proposed universal adversarial example attacks on GCNs. We show that GCNs are highly vulnerable to universal adversarial examples generated by only injecting a single node through experiments.
- 1 February 2020
- Published by IOP Publishing in Journal of Instrumentation
- Vol. 15 (02) , C02011
Current status and plan of development of Nd:YAG laser Thomson scattering system in Heliotron J
An Nd:YAG Thomson Scattering system has been developed in Heliotron J. The Thomson scattering system is used for an investigation of an improved confinement physics such as the edge transport barrier (H-mode) or the internal transport barrier of the helical plasma. The system has 25 spatial points with ~10 mm resolution. Two high repetition Nd:YAG lasers (> 500mJ@50 Hz) realize the measurement of the time evolution of the plasma profile with 10 ms time intervals. The scattered light is collected with a large concave mirror (D=800 mm) with a solid angle of ~80–100 mstr. The laser beam is injected from obliquely downward to upward, and obliquely backscattered light is detected. A multi-path Thomson scattering system is under development to measure anisotropic electron velocity distribution and to increase the accuracy of the density and temperature profiles. First, we have applied a double-path Thomson scattering system using a Faraday rotator to the Heliotron J. The clear backward and forward scattered lights are successfully detected. Now the system is upgraded to a multi-path system using a pockels cell. The final goal of the development is the system that has two pockels cells to separate the backward and forward scattered lights.
- 1 February 2020
- Published by IOP Publishing in Journal of Instrumentation
- Vol. 15 (02) , C02011
Current status and plan of development of Nd:YAG laser Thomson scattering system in Heliotron J
An Nd:YAG Thomson Scattering system has been developed in Heliotron J. The Thomson scattering system is used for an investigation of an improved confinement physics such as the edge transport barrier (H-mode) or the internal transport barrier of the helical plasma. The system has 25 spatial points with ~10 mm resolution. Two high repetition Nd:YAG lasers (> 500mJ@50 Hz) realize the measurement of the time evolution of the plasma profile with 10 ms time intervals. The scattered light is collected with a large concave mirror (D=800 mm) with a solid angle of ~80–100 mstr. The laser beam is injected from obliquely downward to upward, and obliquely backscattered light is detected. A multi-path Thomson scattering system is under development to measure anisotropic electron velocity distribution and to increase the accuracy of the density and temperature profiles. First, we have applied a double-path Thomson scattering system using a Faraday rotator to the Heliotron J. The clear backward and forward scattered lights are successfully detected. Now the system is upgraded to a multi-path system using a pockels cell. The final goal of the development is the system that has two pockels cells to separate the backward and forward scattered lights.