担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Bayesian optimization (BO) improves the efficiency of black-box optimization; however, the associated computational cost and power consumption remain dominant in the application of machine learning methods. This paper proposes a method of determining the stopping time in BO. The proposed criterion is based on the difference between the expectation of the minimum of a variant of the simple regrets before and after evaluating the objective function with a new parameter setting. Unlike existing stopping criteria, the proposed criterion is guaranteed to converge to the theoretically optimal stopping criterion for any choices of arbitrary acquisition functions and threshold values. Moreover, the threshold for the stopping criterion can be determined automatically and adaptively. We experimentally demonstrate that the proposed stopping criterion finds reasonable timing to stop a BO with a small number of evaluations of the objective function.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85165187394&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

DOI： 10.1080/27660400.2022.2140021

DOI： 10.1080/27660400.2022.2140021

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

In this study, we develop a method for multi-task manifold learning. The method aims to improve the performance of manifold learning for multiple tasks, particularly when each task has a small number of samples. Furthermore, the method also aims to generate new samples for new tasks, in addition to new samples for existing tasks. In the proposed method, we use two different types of information transfer: instance transfer and model transfer. For instance transfer, datasets are merged among similar tasks, whereas for model transfer, the manifold models are averaged among similar tasks. For this purpose, the proposed method consists of a set of generative manifold models corresponding to the tasks, which are integrated into a general model of a fiber bundle. We applied the proposed method to artificial datasets and face image sets, and the results showed that the method was able to estimate the manifolds, even for a tiny number of samples.

DOI： 10.1016/j.neucom.2021.11.043

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85121311100&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

The purpose of this study is to develop a method for scalable manifold modeling. A popular method using Gaussian process requires a computational cost of the cubic order for data size, which does not afford to apply large-scale datasets. We aim to achieve the linear order in computational cost using unsupervised kernel regression and its sparse approximation instead of using the Gaussian process regression. For this purpose, we introduce two types of sparse approximations; one is the discretization of the latent space by a grid of inducing points, and the other is a sparse matrix decomposition of the local and global kernel matrices. We evaluated the computation time of the proposed method by using an artificial dataset, and the results showed that the proposed method achieved the linear order in computation time.

DOI： 10.1109/SCISISIS55246.2022.10001865

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85146734984&origin=inward

記述言語：日本語   掲載種別：研究論文（研究会，シンポジウム資料等）

<p>【背景】深層学習によって特徴量設計からニューラルネットの構造設計にパラダイムシフトが起こったが, タスクに応じた構造設計は難しい. そこでニューラルネットの構造の自動探索手法―ニューラルアーキテクチャサーチ (<i>Neural architecture search</i>: NAS)―が盛んに研究されている. 最新手法では低コストでの構造探索に成功している. しかし, 1) 精度と探索コストにはトレードオフがある. 2) 性能劣化を防ぐための早期停止が有効である. ということが報告されており, いつ探索を停止するか？という課題が存在する. 【目的】NASの最適停止手法を提案する. 性能の良い構造を効率よく探索することを目的とする. 【結果】数値実験により提案手法が有効であることを検証した.</p>

DOI： 10.11517/pjsai.jsai2022.0_3j4os3b01

CiNii Research

記述言語：日本語   掲載種別：研究論文（研究会，シンポジウム資料等）

<p>近年のマテリアルズ・インフォマティクスの興隆にともなって、物質科学における様々な実験の効率化、自動化・自律化が望まれている。我々は物質科学の主要な実験手法のひとつであるスペクトル測定を、能動学習を用いて効率化する手法を開発した。獲得関数が最大となるエネルギー点を順次計測していく能動学習によって、実験者が介在することなく自動かつ最適な条件でのスペクトル測定を実現した。ガウス過程回帰の期待汎化誤差の上界に基づく停止基準を用いることで、スペクトルの種類に依らず測定の自動停止が可能になった。これによって従来のスペクトル計測に対して少数の計測点で同等の物質情報を得ることができる。</p>

DOI： 10.11517/pjsai.jsai2022.0_3yin208

CiNii Research

記述言語：英語   掲載種別：研究論文（学術雑誌）

The automated stopping of a spectral measurement with active learning is proposed. The optimal stopping of the measurement is realised with a stopping criterion based on the upper bound of the posterior average of the generalisation error of the Gaussian process regression. It is revealed that the automated stopping criterion of the spectral measurement gives an approximated X-ray absorption spectrum with sufficient accuracy and reduced data size. The proposed method is not only a proof-of-concept of the optimal stopping problem in active learning but also the key to enhancing the efficiency of spectral measurements for high-throughput experiments in the era of materials informatics.

DOI： 10.1038/s41524-021-00606-5

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85113396806&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Visual analytics (VA) is a visually assisted exploratory analysis approach in which knowledge discovery is executed interactively between the user and system in a human-centered manner. The purpose of this study is to develop a method for the VA of set data aimed at supporting knowledge discovery and member selection. A typical target application is a visual support system for team analysis and member selection, by which users can analyze past teams and examine candidate lineups for new teams. Because there are several difficulties, such as the combinatorial explosion problem, developing a VA system of set data is challenging. In this study, we first define the requirements that the target system should satisfy and clarify the accompanying challenges. Then we propose a method for the VA of set data, which satisfies the requirements. The key idea is to model the generation process of sets and their outputs using a manifold network model. The proposed method visualizes the relevant factors as a set of topographic maps on which various information is visualized. Furthermore, using the topographic maps as a bidirectional interface, users can indicate their targets of interest in the system on these maps. We demonstrate the proposed method by applying it to basketball teams, and compare with a benchmark system for outcome prediction and lineup reconstruction tasks. Because the method can be adapted to individual application cases by extending the network structure, it can be a general method by which practical systems can be built.

DOI： 10.1016/j.dss.2021.113635

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85111295128&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In this paper, we propose a method for multi-task manifold learning. For a set of tasks of dimensionality reduction, the aim of the method is to model each given dataset as a manifold, and map it to a low-dimensional space. For this purpose, we use a hierarchical manifold modeling approach. Thus, while each data distribution is represented by a manifold model, the obtained models are further modeled by a higher-order manifold in a function space. The higher-order model mediates the information transfer between tasks, and as a result, the performance of each task is improved. The results of simulations show that the proposed method can estimate manifolds approximately, even in cases in which a tiny number of samples are provided for each task.

DOI： 10.1007/978-3-030-04182-3_34

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85059020589&origin=inward

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

The purpose of this study is to propose a paradigm visualizing the viewpoints from datasets observed by multiple viewpoints, which is referred to as Latent Viewpoint Visualization (LVV). Since LVV visualizes similarity/dissimilarity among the viewpoints, it has many applications such as the authors’ perspective from news articles and the psychological measurements’ aspect from psychological surveys. In this study, we propose the concept of LVV and develop a preliminary algorithm. Furthermore, we experimentally show what kind of information can be visualized by LVV using several datasets.

DOI： 10.1007/978-3-030-04182-3_56

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85059046342&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

The aim of this work is to develop a visualization method for multilevel–multigroup analysis based on a multiway nonlinear dimensionality reduction. The task of the method is to visualize what kinds of members each group is composed and to visualize the similarity between the groups in terms of probability distribution of constituent members. To achieve the task, the proposed method consists of hierarchically coupled tensor self-organizing maps, corresponding to the member/group level. This architecture enables more flexible analysis than ordinary parametric multilevel analysis, as it retains a high level of interpretability supported by strong visualization. We applied the proposed method to one benchmark dataset and two practical datasets: one is the survey data on the football players belonging to different teams and the other is the employee survey data belonging to different departments in a company. Our method successfully visualizes the types of the members that constitute each group as well as visualizes the differences or similarities between the groups.

DOI： 10.1007/s11063-017-9643-1

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85020251157&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

<p>The aim of this work is to develop a method for the simultaneous analysis of multiple groups and their members based on hierarchical tensor manifold modeling. The method is particularly designed to analyze multiple teams, such as sports teams and business teams. The proposed method represents members' data using a nonlinear manifold for each team, and then these manifolds are further modeled using another nonlinear manifold in the model space. For this purpose, the method estimates the role of each member in the team, and discovers correspondences between members that play similar roles in different teams. The proposed method was applied to basketball league data, and it demonstrated the ability of knowledge discovery from players' statistics. We also demonstrated that the method could be used as a general tool for multi-level multi-group analysis by applying it to marketing data.</p>

DOI： 10.1587/transfun.E101.A.1745

CiNii Article

その他リンク： https://ci.nii.ac.jp/naid/130007502924

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Kyutacar

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

© 2016 IEEE. In this paper, we describe two types of selforganizing maps (SOMs): a self-organizing system map (SOSM) and a self-organizing view map (SOVM). The task of SOSM is to discover meta-knowledge by estimating the latent system parameter from a series of experiments. The SOSM algorithm is almost identical to the 'SOM of SOMs' (SOM2), with a small modification. SOVM aims to discover meta-knowledge by estimating the latent view parameter. SOSM enables higher-order data structures to be estimated, whereas SOVM enables the consideration of heterogeneous datasets. Simulation results show that these algorithms work as expected.

DOI： 10.1109/SCIS-ISIS.2016.0078

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85010366338&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This paper describes a method of multilevel–multigroup analysis based on a nonlinear multiway dimensionality reduction. To analyze a set of groups in terms of the probabilistic distribution of their constituent member data, the proposed method uses a hierarchical pair of tensor self-organizing maps (TSOMs), one for the member analysis and the other for the group analysis. This architecture enables more flexible analysis than ordinary parametric multilevel analysis, as it retains a high level of translatability supported by strong visualization. Furthermore, this architecture provides a consistent and seamless computation method for multilevel–multigroup analysis by integrating two different levels into a hierarchical tensor SOM network. The proposed method is applied to a dataset of football teams in a university league, and successfully visualizes the types of players that constitute each team as well as the differences or similarities between the teams.

DOI： 10.1007/978-3-319-46675-0_50

Kyutacar

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84992745797&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Kyutacar