|
Affiliation Department |
Department of Computer Science
|
|
Title |
Professor |
|
Contact information |
|
|
External Link |
|
INUZUKA Nobuhiro
|
|
|
Degree
-
Doctor (Engineering) ( Nagoya Institute of Technology )
-
Master of Engineering ( Nagoya Institute of Technology )
Research Areas
-
Informatics / Intelligent informatics
-
Informatics / Database
-
Informatics / Theory of informatics
From School
-
Nagoya Institute of Technology Faculty of Engineering Graduated
- 1987.03
Country：Japan
From Graduate School
-
Nagoya Institute of Technology Graduate School, Division of Engineering Doctor's Course Completed
- 1992.03
Country：Japan
External Career
-
Research Fellow of the Japan Society for the Promotion of Science Special researcher of the Japan Society for the Promotion of Science
1990.04 - 1992.03
Country：Japan
-
Japan Society for the Promotion of Science, Postdoctoral Fellow for Research Abroad Special researcher of the Japan Society for the Promotion of Science
1994.04 - 1996.03
Country：Japan
Professional Memberships
-
The Japanese Association of Student Counseling
2012.01
-
Philosophy of Science Society, Japan
2008.04
-
Association for Computing Machinery
2005.04
-
The Society if Instrument and Control Engineers
2004.04
-
Information Processing Society of Japan
2003.04
Research Career
-
Knowledge Discovery Algorithms for Analyzing Human Relationships
Grant-in-Aid for Scientific Research
Project Year： 2015.04 - 2019.03
-
Development of Student Counseling Record Management System Forcusing Counseling Structure
The Other Research Programs
Project Year： 2011.12 - 2012.07
-
Research in Information Technology to Establish Scientific and Reasonable Methodology in Student Counseling
Grant-in-Aid for Scientific Research
Project Year： 2011.05 - 2013.03
-
Usability Tests of a Multi-Relational Data Mining Method
JST Comprehensive Support Programs for Creation of Regional Innovation
Project Year： 2010.10 - 2011.03
-
Development of a Data Analysis System Based-on Relation Data Mining Methods
JST Comprehensive Support Programs for Creation of Regional Innovation
Project Year： 2009.10 - 2010.03
Papers
-
Store Cluster Analysis by Extracting Order Trends Using POS Data Reviewed
2023-BIO-73 ( 12 ) 1 - 6 2023.10
Authorship：Last author Language：Japanese Publishing type：Research paper (scientific journal)
POS data analysis is a technique for analyzing consumer purchasing behavior. In order to improve product sales, there is a technique for clustering stores based on the ratio of product sales by genre to total retail store sales, and clarifying the relationship between location and the tendency to order at stores within a cluster. However, in restaurants, the trend is not in the number of product sales, but in the content of individual customer orders. In this study, we propose a method to classify stores by extracting ordering trends from the orders of each receipt using non-negative matrix factorization. Since this method classifies stores based on frequently occurring order trends, it can be used to identify the relationship between store needs and location/environment. In an experiment, we applied the proposed method to POS data of several restaurants and confirmed its effectiveness.
-
Effect of Curiosity-driven Search in Multi-agent Deep Reinforcement Learning Reviewed
16 ( 2 ) 80 - 90 2023.10
Authorship：Last author Language：Japanese
In a multi-agent environment where multiple agents exist, the agents may achieve better results by cooperating one another than by acting self-interestedly. Curiosity-driven exploration, which is used in a single-agent environment with sparse rewards, offers rewards to unknown states, and the agent learns a policy to seek such unknown states accordingly. Using it in a multi-agent environment may allow the agents to acquire cooperative behavior by promoting exploration and changing the reward structure. The experiments we conducted in a predator-prey problem show that the curiosity-driven exploration allows the predators to catch more preys.
-
Evacuation Planning Model Considering Diurnal Variation in the Number of People Staying
16 ( 2 ) 103 - 109 2023.10
Authorship：Last author Language：Japanese Publishing type：Research paper (scientific journal)
In recent years, we have seen many large-scale disasters such as earthquakes, and overflows of shelter capacity often occurred due to the day-night population gap at the time of the Great East Japan Earthquake. Therefore, it has become an important issue to allocate shelters in consideration of the day-night population gap (diurnal variation). In the shelter assignment problem, the use of zero-suppressed dichotomous decision graphs (ZDD) makes it possible to enumerate all patterns of shelter assignment. Studies using ZDD evaluated the distance to the shelter and the capacity factor, but did not consider the dwell time on the travel route, and used the fixed number of evacuees. In this study, we add an evaluation index using a network flow simulation method in order to take into account the number of people who stay in shelters along travel routes. In addition, we propose a method for obtaining shelter allocation that takes into account the diurnal variation in the number of people staying in shelters using stay history data for all time segments. In our experiments, we used three evaluation values: distance, occupancy rate, and evacuation completion time, and confirmed the effectiveness of the method through experiments using real environments.
-
An Analysis of Stylistic Feature Changes in Language Use with Conversation Volume on Twitter Reviewed
16 ( 1 ) 1 - 7 2023.03
Authorship：Last author Language：Japanese Publishing type：Research paper (scientific journal)
In the field of sociolinguistics, it has been shown that people manipulate their social distance from others and increase the efficiency of communication by changing their language in conversation based on accommodation theory. In social psychology, the mere exposure effect, in which repeated contact improves the impression the other person, has been identified, and some studies have shown that the greater the number of contacts in a short period, the better the impression. Conversation is one way people make contact, and we can assume that people manipulate their relationships with others by improving their impressions of others and changing their language through increased conversation. On the other hand, the platform Twitter enables the observation of people's behaviour on a large scale. `Reply' on Twitter can be regarded as conversations with others. In this study, we propose a model to quantitatively extract changes in stylistic features of language with the volume of conversation per unit time on Twitter. In the proposed model, we convert the stylistic features of the conversation data into a textual feature matrix. Then, we analyze the conversation data by extracting important textual features that increase or decrease with the conversation volume per unit time using non-negative matrix factorization. Experimental results showed a number of linguistic findings, such as the fact that increased volume of conversations leads to fewer polite expressions and more straightforward sentences.
Other Link： https://ipsj.ixsq.nii.ac.jp/ej/?action=repository_uri&item_id=225568&file_id=1&file_no=1
-
隠れインデックスファンド検出のための One-class SVM 出力校正法
籔内 陽斗, 松井 藤五郎, 武藤 敦子, 島 孔介, 森山 甲一, 犬塚 信博
研究報告バイオ情報学（BIO） 2023 ( FIN-030 ) 32 - 39 2023.03
Authorship：Last author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
ファンド(投資信託)の中には目論見書に記された運用方針と実際の運用方針が異なる場合がある。先行研究では実際の運用方針に基づくファンドのグループ分けを目的に、月次リターンの時系列に対してk-meansとUMAPを組み合わせた手法を用いてファンドをクラスタリングする方法が提案された。しかし、k-meansによるクラスタリングではラベル情報を一切使用しないという問題がある。本論文では、インデックスファンドのラベルは正しいと仮定し、ラベル情報を使用可能な手法であるOne-class SVMを導入する。ところが、通常のOne-class SVMでは、ラベルが付与されたインデックスファンドだけをインデックスと判別するモデルを作成するため、インデックスに類似した隠れインデックスファンドをインデックスと判別することができない。この問題を解決するために、本論文ではOne-class SVMモデルの出力値に着目し、One-class SVMの出力を校正する新しい方法を提案する。提案手法を同一の指標をベンチマークとするファンド群に対して適用することで隠れインデックスファンドを見つけることができた。
-
マルチエージェント深層強化学習における好奇心探索の影響
岩科 亨, 森山 甲一, 松井 藤五郎, 武藤 敦子, 犬塚 信博
研究報告数理モデル化と問題解決（MPS） 2023-MPS-142 ( 14 ) 1 - 6 2023.03
Authorship：Last author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
マルチエージェント環境には複数のエージェントが存在するが，それぞれが利己的な行動を獲得してしまうよりも，協力することでより良い結果が得られる可能性がある．報酬が疎なシングルエージェント環境で利用される好奇心探索は，未知の状態に対し報酬を与えることで最大化すべき報酬が変化し，それに伴い方策も未知の状態を求めるよう変化する．それをマルチエージェント環境で用いることで，探索拡大に加え，報酬の構造の変化により，獲得できていなかった協力的な行動に繋がる可能性がある．好奇心探索を用いて追跡問題で実験したところ，より多くの獲物を捕まえることができた．
-
滞在人数の日内変動を考慮した避難計画モデル
山本 正也, 武藤 敦子, 島 孔介, 森山 甲一, 松井 藤五郎, 犬塚 信博
研究報告バイオ情報学（BIO） 2023-BIO-73 ( 25 ) 1 - 5 2023.03
Authorship：Last author Language：Japanese Publishing type：Research paper (conference, symposium, etc.)
近年，地震などの大規模な災害の発生が多く見られ，東日本大震災の際には，昼夜人口のギャップにより容量溢れが多く発生した．そのため，昼夜人口のギャップ（日内変動）を考慮して避難所を割り当てることは重要な課題になっている．避難所割り当て問題において，ゼロサプレス型二分決定グラフ（ZDD）を用いることで避難所割り当てのパターンを全列挙することが可能になっている．ZDD を用いた研究では，避難所までの距離と収容率で評価をしているが，移動経路での滞留は考慮しておらず，避難対象者は常に固定した人数で行っている．そこで本研究では，移動経路での滞留を考慮するためにネットワークフローによるシミュレーション手法での評価を追加し，滞在履歴データを用いて全時間区分で評価を行うことで滞在人数の日内変動を考慮した避難所割り当てを求める手法を提案する．距離・収容率・避難完了時間の 3 つを評価値とし，実環境を用いた実験により効果を確認した．
-
Evaluation of Employees by Generating Meeting Networks Using Entry and Exit Data Reviewed
Atsuko Mutoh, Kazuma Morikida, Koichi Moriyama, Nobuhiro Inuzuka
Proceedings of 2023 IEEE International Conference on Consumer Electronics (ICCE) 1 - 4 2023.01
Language：English Publishing type：Research paper (international conference proceedings)
Data related to various human movements are available due to the recent development of IoT devices and communication networks, and research on the use of historical data of human movements is being actively performed. In this study, we use log data on employee movement obtained from an entry/exit management system, which has been implemented in many businesses. We propose a method for creating a social network based on employee entry/exit history and observing employee performance. We focus on meeting room-related data from entry/exit data and evaluate employees using meeting network centrality indexes. According to the experimental results, the meeting network's closeness centrality best represents job position, while the meeting network's betweenness centrality best represents employee performance.
DOI： 10.1109/ICCE56470.2023.10043416
Other Link： https://ieeexplore.ieee.org/abstract/document/10043416
-
Evaluation System for Martial Arts Demonstration from Smartphone Sensor Data Using Deep Neural Networks on Noisy Labels Reviewed International journal
Shohei Yamanaka, Kosuke Shima, Atsuko Mutoh, Koichi Moriyama, Tohgoroh Matsui, Nobuhiro Inuzuka
Proceedings of 2023 IEEE International Conference on Consumer Electronics (ICCE) 1 - 5 2023.01
Authorship：Last author Language：English Publishing type：Research paper (international conference proceedings)
Smartphones and smartwatches are widespread. Devices have with sensors. Research activities to obtain data from sensors to recognize sports motion and find modifications are on the rise. These studies make it possible to provide movement feedback without a coach. However, a few studies have focused on the correctness of the movement, or which parts of the movement are correct. Herein, we propose a method to evaluate the movements of a martial arts demonstration based on sensor data obtained from a smartphone. In the demonstration, several movements, such as punching and kicking, are performed. The coach considers each movement and evaluates the overall movement comprehensively. Regarding the coach evaluation, we evaluate sensor data by dividing the data and using machine learning. This method was applied to the acceleration data of taekwondo demonstrations. It was shown to reproduce the instructor's score with high accuracy.
DOI： 10.1109/ICCE56470.2023.10043511
Other Link： https://ieeexplore.ieee.org/abstract/document/10043511
-
Extraction of Behavioral Patterns by Recombining Non-Negative Multiple Matrix Factorization and Clustering Results Reviewed International journal
Shogo Yasui, Atsuko Mutoh, Koichi Moriyama, Tohgoroh Matsui, and Nobuhiro Inuzuka
Proceedings of Proceedings of 2021 IEEE 10th Global Conference on Consumer Electronics (GCCE 2021) 2021.10
Language：English Publishing type：Research paper (international conference proceedings) Publisher：IEEE
Books and Other Publications
-
MIT認知科学大事典
中島秀之, 他（ Role： Joint translator）
共立出版 2012.11 （ ISBN:978-4-320-09447-5 ）
Language：jpn Book type：Dictionary, encyclopedia
「認知科学」の全分野にわたって，それぞれの方法論および理論を網羅した世界に類のない事典。認知科学を構成している六つの主要分野：哲学，心理学，神経科学，計算論的知能，言語学，文化・認知・進化の中から470項目を厳選し，それぞれに対して第一級の研究者が執筆にあたっている。
登録者（犬塚）は人工知能関係の８項目程度を翻訳した。
Presentations
-
エージェント間の距離がタスク達成に影響する環境下における報酬の制御
中田瑛，森山甲一，武藤敦子，松井藤五郎，犬塚信博
エージェント間の距離がタスク達成に影響する環境下における報酬の制御 SMASH21 SUMMER SYMPOSIUM
Event date： 2021.09
Language：Japanese Presentation type：Oral presentation (general)
Venue：ソフトウェア科学会
-
Curiosity-Driven Search in a Multiagent Reinforcement Learning Problem
Toru Iwashina, Koichi Moriyama, Tohgoroh Matsui, Atsuko Mutoh, Nobuhiro Inuzuka
The 35th Annual Conference of the Japanese Society for Artificial Intelligence Japanese Society for Artificial Intelligence
Event date： 2021.06
Language：English Presentation type：Oral presentation (general)
Reinforcement learning has been shown to be capable of
dealing with complex control problems, e.g., automated driving. In the real world, on the other hand, people are not alone but interacting with one another. When we apply reinforcement learning to such a multiagent environment, an agent should find a way to get along with others. To do so, the agent needs to search in its strategy space efficiently. In this work, we apply the curiosity-driven search to a multiagent environment. The curiosity-driven search, which is proposed for single agent scenarios, generates intrinsic rewards that guide the agent in sparse-reward environments. The intrinsic rewards may give a good result also in a multiagent environment where the agents need to find a good combination of actions. We conducted experiments using a predator-prey problem popular in multiagent studies, and found that the learning speed in the early stage of learning was improved compared to the case without the curiosity-driven search. -
Evaluation of martial arts demonstration focusing on motion timing using acceleration data
Shohei Yamanaka, Kousuke Shima, Atsuko Mutoh, Koichi Moriyama, Tohgoroh Matsui, Nobuhiro Inuzuka
The 35th Annual Conference of the Japanese Society for Artificial Intelligence Japanese Society for Artificial Intelligence
Event date： 2021.06
Language：Japanese Presentation type：Oral presentation (general)
In martial arts such as taekwondo, it is important to know one's own skill level in order to improve. In this study, we propose a method to evaluate the performance of taekwondo, attaching a smartphone with the waist and collect acceleration data. In martial arts, including taekwondo, when a martial artist stops moving, he or she instantly holds the entire body still. The acceleration data shows a peak at the moment of this stillness. If the timing of this peak is the same in all three axes, the fighter is considered to be more skilled. In this paper, We use the difference in the timing of the three axes to calculate the feature value, and decision trees create to evaluate the movement from the acceleration data using these features as explanatory variables. We evaluated the effectiveness of decision trees and confirmed these accuracy.
-
Investment Trusts Clustering Using UMAP For The Long Return International conference
Haruto Yabuuchi, Tohgoroh Matsui, Atsuko Mutoh, Koichi Moriyama, Nobuhiro Inuzuka
The 35th Annual Conference of the Japanese Society for Artificial Intelligence Japanese Society for Artificial Intelligence
Event date： 2021.06
Language：Japanese Presentation type：Oral presentation (general)
It is necessary to compare investment trusts with their investment similarity, because their actual operations
may differ from their investment policies. A method for clustering investment trusts based on their monthly return has been proposed, but it only uses 3months. In this paper, we propose a method for clustering investment trusts with long term return and using the UMAP. -
Optimization of evacuation shelter allocation avoiding intersections of evacuation routes by network division
Masaya Yamamoto, Atsuko Mutoh, Koichi Moriyama, Tohgoroh Matsui, Nobuhiro Inuzuka
The 35th Annual Conference of the Japanese Society for Artificial Intelligence Japanese Society for Artificial Intelligence
Event date： 2021.06
Language：Japanese Presentation type：Oral presentation (general)
In recent years, there have been many occurrences of earthquakes and other large-scale disasters. Since the Great East Japan Earthquake revealed that many residents do not know where to evacuate to in the event of a large-scale disaster, evacuation shelter allocation is currently an important issue. Okada et al. proposed a method for allocating evacuation centers using data on people's stay histories. However, this method has a problem that it does not take into account evacuation routes. In this study, we propose a method to determine whether or not the network in the area to be evacuated can be divided into buildings and evacuation shelters, and to obtain the allocation of evacuation shelters without intersecting evacuation routes. As a result of evaluation experiments, we found that the proposed method was able to avoid the crossing of evacuation routes, while the conventional method had the possibility of crossing evacuation routes.
-
社会ネットワークの社会的属性を特徴づける構造的属性に関する研究 International conference
細木 聡, 犬塚 信博, 武藤 敦子, 森山 甲一
人工知能学会第35回全国大会 人工知能学会
Event date： 2021.06
Language：Japanese Presentation type：Oral presentation (general)
社会ネットワーク分析において、各ノードの周囲の構造的属性は有用な働きをする可能性がある。夏目らは、構造的異質性属性を与えるオペレータを定義し、これによって得られた属性を適用した形式概念分析によって、具体的なエゴの振る舞いを分析した。しかし、限られた属性とオペレータの適用方法を対象としており、構造と関連する広い範囲の属性を対象とする必要がある。そこで、本研究では構造的な異質性に基づく新たな属性を定義し、それらも含めて有用性を調査した。その結果、いくつかのオペレータと属性の組合せで有用な結果を得た。
-
Estimation of performance evaluation factors by analyzing the centrality in the meeting network International conference
Yuta Mizuno, Atsuko Mutoh, Koichi Moriyama, Tohgoroh Matsui, Nobuhiro Inuzuka
The 35th Annual Conference of the Japanese Society for Artificial Intelligence Japanese Society for Artificial Intelligence
Event date： 2021.06
Language：Japanese Presentation type：Oral presentation (general)
In recent years, some researches aimed at searching for important persons on social networks using the centrality have been focused. In this study, by measuring the correlation between each centrality of the meeting network in the office and various indicators, we clarified the characteristics of each centrality, and estimated the factors for evaluating the performance in the company.
-
非負値多重行列因子分解による購買および TV視聴パターンの可視化
安井彰悟, 武藤敦子, 森山甲一, 松井藤五郎, 犬塚信博
情報処理学会第83回全国大会 情報処理学会
Event date： 2021.03
Language：Japanese Presentation type：Oral presentation (general)
本研究では，小島らの非負値多重行列因子分解とクラスタリングを用いたユーザ行動パターン抽出方法における，因子分解結果からクラスタ内ユーザの行動特徴を解釈するのが複雑であるという問題に対して，因子行列と各クラスタ中心のユーザの因子寄与度を掛け合わせることによって，クラスタ内ユーザの行動特徴を可視化する手法を提案する。
その後，従来手法と同様に，クラスタ内ユーザの個人属性を決定木学習することで，行動特徴と個人属性の関係性について分析する。
最後に，ユーザの商品の購買と，TV視聴時間帯の関係性について本手法を用いて分析を行い，有効性を確認する。 -
Automatic Generation of Behavioral Rules for Pedestrian Simulations Using Reinforcement Learning
Himeka Kobayashi, Koichi Moriyama, Tohgoroh Matsui, Atsuko Mutoh, Nobuhiro Inuzuka
Event date： 2021.03
Language：Japanese Presentation type：Oral presentation (general)
In designing pedestrian simulations, it is common for model designers to prepare in advance behavioral rules of the pedestrians. To reduce the burden on the designers, we aim to automatically generate the rules using reinforcement learning. In this work, we consider counter-flow simulations where agents move toward each other. They decide their actions based on visual information. This work designs their states to avoid collisions in the simulations. From the simulation experiments, we found that the proposed agents learn appropriate behavioral rules to arrive at the destination while avoiding collisions.
-
高次元連続観測空間における安全な強化学習
梅本 匠, 松井 藤五郎, 武藤 敦子, 森山 甲一, 犬塚 信博
情報論的学習理論と機械学習研究会 電子情報通信学会
Event date： 2021.03
Language：Japanese Presentation type：Oral presentation (general)
Venue：オンライン
本論文では，連続状態空間における成功確率と収益に基づく安全な強化学習の手法であるCSEQを高次元に拡張する方法を提案する．被災地や宇宙などの人が直接行くことのできない危険な環境でロボットの活躍が期待されている．強化学習は試行錯誤に基づいてより良い行動を学習する機械学習の手法であり，強化学習手法の中でも危険回避行動を学習することに着目した安全な強化学習として成功確率と収益に基づくEQという手法が提案されている．これを連続状態空間に拡張した手法がCSEQであり，2次元空間上のシンプルな問題ではその有効性が確認されている．しかしながら，安全な強化学習を活用したいロボットの問題などの観測値は高次元で与えられることがほとんどである．そこで我々はVAEでモデル化された潜在変数の平均値を用いて高次元連続空間を扱うことのできる安全な強化学習を提案する．また，高次元連続状態空間の例題を用いてその有効性を検証した．
Industrial Property Rights
-
Record Management System Conseling based-on Counseling Structure
Nobuhiro Inzuuka, Atsuko Mutoh, Naotaka Oda
Application no：特願2011-999999 Date applied：2011.02
Country of applicant：Domestic Country of acquisition：Domestic
-
Friendship Prediction System
Nobuhiro Inuzuka, Tomofumi Nakano, Kosaku Shimomura, Hiroshi Matsuo
Application no：特願2008-34078 Date applied：2008.02
Announcement no：特開2009-193396 Date announced：2009.08
Patent/Registration no：5283059 Date registered：2013.06 Date issued：2013.06
Country of applicant：Domestic Country of acquisition：Domestic
Awards
-
2019年度人工知能学会全国大会優秀賞
2019.06 人工知能学会 非負値多重行列因子分解の因子行列を用いたクラスタリングと決定木学習によるオフィスの入退室データの分析
小島 世大, 石槫 隼人, 坂田 美和, 武藤 敦子, 森山 甲一, 犬塚 信博
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
情報処理学会全国大会学生奨励賞
2013.03 情報処理学会
西山瑞紀
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
第９回情報学ワークショップ奨励賞
2011.11 第９回情報学ワークショップ実行委員会
小田尚宜, 武藤敦子, 犬塚信博
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
第８回情報学ワークショップ奨励賞
2010.12 情報学ワークショップ実行委員会
松島裕, 犬塚信博
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
第８回情報学ワークショップ優秀論文賞
2010.12 情報学ワークショップ実行委員会
中野裕介, 犬塚信博
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
第７回情報学ワークショップ奨励賞
2009.11 情報学ワークショップ実行委員会
牧野敏行,犬塚信博
Award type：Award from Japanese society, conference, symposium, etc. Country：Japan
-
電気関係学会東海支部連合大会奨励賞
1996.04 -
-
Country：Japan
Scientific Research Funds Acquisition Results
-
人間関係分析のための知識発見アルゴリズム
2015.04 - 2019.03
科学研究費補助金 基盤研究(C)
犬塚信博
-
科学的・合理的な学生相談手法を確立するための情報科学的研究
2011.04 - 2013.03
科学研究費補助金 挑戦的萌芽研究
犬塚信博、粥川 裕平, 武藤 敦子