English
広島市立大学 
情報科学研究科 
知能工学専攻 

教授 
高濱 徹行 
タカハマ テツユキ 
Takahama Tetsuyuki 

Tel.082-830-1698  
 
個人ウェブサイトはこちら  

学歴
京都大学  工学部  電気工学第二学科  1982  卒業 
京都大学  工学研究科  電気工学第二専攻  博士  1987  単位取得満期退学 

学位
工学修士  京都大学 
博士(工学)  京都大学 

研究分野
計算知能 
進化的計算 
最適化 

研究キーワード
ナチュラル・コンピューティング 
進化的アルゴリズム 
最適化アルゴリズム 

研究テーマ
ファジィ制御ルールの学習アルゴリズムに関する研究  1990-現在 
多目的最適化手法に関する研究  1996-現在 
α制約法による制約付き最適化問題の解法  1997-現在 
退化による構造学習に関する研究  1999-現在 

著書
"A Study on Selecting an Oblique Coordinate System for Rotation-Invariant Blend Crossover in a Real-Coded Genetic Algorithm", in Recent Studies in Economic Sciences: Information Systems, Project Managements, Economics, OR and Mathematics  Setsuko Sakai and Tetsuyuki Takahama  65-87  Kyushu University Press  2018/03  Differential Evolution (DE) has been successfully applied to various optimization problems. The performance of DE is affected by algorithm parameters such as a scaling factor F and a crossover rate CR. Many studies have been done to control the parameters adaptively. One of the most successful studies on controlling the parameters is JADE. In JADE, the values of each parameter are generated according to one probability density function (PDF) which is learned by the values in success cases where the child is better than the parent. However, search performance might be improved by learning multiple PDFs for each parameter based on some characteristics of search points. In this study, search points are divided into plural groups according to some criteria and PDFs are learned by parameter values in success cases for each group. Objective values and distances from a reference point, which is the best search point or the centroid of search points, are adopted as the criteria. The effect of JADE with group-based learning is shown by solving thirteen benchmark problems. 
"A Comparative Study on Grouping Methods for an Adaptive Differential Evolution", in Challenging Researches in Economic Sciences: Legal Informatics, Environmental Economics, Economics, OR and Mathematics  Setsuko Sakai and Tetsuyuki Takahama  51-91  Kyushu University Press  2017/03  Differential Evolution (DE) has been successfully applied to various optimization problems. The performance of DE is affected by algorithm parameters such as a scaling factor F and a crossover rate CR. Many studies have been done to control the parameters adaptively. One of the most successful studies on controlling the parameters is JADE. In JADE, the values of each parameter are generated according to one probability density function (PDF) which is learned by the values in success cases where the child is better than the parent. However, search performance might be improved by learning multiple PDFs for each parameter based on some characteristics of search points. In this study, search points are divided into plural groups according to some criteria and PDFs are learned by parameter values in success cases for each group. Objective values and distances from a reference point, which is the best search point or the centroid of search points, are adopted as the criteria. The effect of JADE with group-based learning is shown by solving thirteen benchmark problems. 
"A Comparative Study on Detecting Ridge Structure for Population-Based Optimization Algorithms", in Contemporary Works in Economic Sciences: Legal Informatics, Economics, OR and Mathematics  Setsuko Sakai, Tetsuyuki Takahama  61-82  Kyushu University Press  2016/02 
"A Study on Adaptive Parameter Control for Interactive Differential Evolution Using Pairwise Comparison", in New Solutions in Legal Informatics, Economic Sciences and Mathematics  Setsuko Sakai, Tetsuyuki Takahama  101-121  Kyushu University Press  2015/03 
"A Comparative Study on Estimation Methods of Landscape Modality for Evolutionary Algorithms" in Legal Informatics, Economic Science and Mathematical Research  Setsuko Sakai, Tetsuyuki Takahama  55-80  Kyushu University Press  2014/03 
詳細表示...

論文
研究論文(学術雑誌)  共著  The Velocity Updating Rule According to an Oblique Coordinate System with Mutation and Dynamic Scaling for Particle Swarm Optimization  T.Takahama, S.Sakai  Artificial Life and Robotics  23, (to appear)-  2018  Particle swarm optimization (PSO) has been showing powerful search performance especially in separable and unimodal problems. However, the performance is deteriorated in non-separable problems such as rotated problems. In this study, a new velocity updating rule according to an oblique coordinate system, instead of an orthogonal coordinate system, is proposed to solve non-separable problems. Two mutation operations for the best particle and the worst particle are proposed to improve the diversity of particles and to decrease the degradation of moving speed of particles. Also, the vectors generated according to the oblique coordinate system is dynamically scaled in order to improve the robustness and efficiency of the search. The advantage of the proposed method is shown by solving various problems including separable, non-separable, unimodal, and multimodal problems, and their rotated problems and by comparing the results of the proposed method with those of standard PSO. 
研究論文(学術雑誌)  共著  Particle swarm optimisation with dynamic search strategies based on rank correlation  T.Nishio, J.Kushida, A.Hara, T.Takahama  International Journal of Computational Intelligence Studies  6/ 4, 311-332  2018/01 
研究論文(国際会議プロシーディングス)  共著  Particle Swarm Optimization with the Velocity Updating Rule According to an Oblique Coordinate System  Tetsuyuki Takahama and Setsuko Sakai  Proc. of the 2nd International Symposium on Swarm Behavior and Bio-Inspired Robotics (SWARM2017)  70-77  2017/10  Optimization problems have some characteristics: Dependency among decision variables such as separable or non-separable problems, and landscape modality such as unimodal or multimodal problems. Particle swarm optimization (PSO) has been shown powerful search performance especially in separable and unimodal problems. However, the performance is deteriorated in non-separable problems such as rotated problems. Although velocity updating rules using random rotation matrices have been proposed to solve non-separable problems, the computational cost of generating the random rotation matrices is very high. In this study, a new velocity updating rule according to an oblique coordinate system, instead of an orthogonal coordinate system, is proposed to solve non-separable problems. Also, two mutation operations for the worst particle and the best particle are proposed to improve the diversity and the convergence of particles, respectively. The advantage of the proposed method is shown by solving various problems including unimodal problems, multimodal problems, and rotated problems and by comparing the results of the proposed method with those of standard PSO. 
研究論文(国際会議プロシーディングス)  共著  An Adaptive Differential Evolution with Exploitation and Exploration by Extreme Individuals  Tetsuyuki Takahama and Setsuko Sakai  Proc. of SICE Annual Conference 2017 (SICE2017)  1147-1152  2017/09  In a natural population, extreme individuals are very important for survival of the population. When the main population is destroyed by catastrophes, the few extreme individuals gain significance and will insure the survival of the population. Differential Evolution (DE) has been successfully applied to various optimization problems. However, DE sometimes trapped into some local solutions, which is a kind of a catastrophe. In this study, the extreme individuals are paid attention to. The best individuals perform exploitation or local search to keep extremely good positions. The worst individuals perform exploration or global search to keep positions far from the best individuals. Other individuals perform adaptive search based on JADE which is one of the most successful algorithms on controlling algorithm parameters. In JADE, the values of two algorithm parameters are generated according to two probability density functions which are learned by the values in success cases where the child is better than the parent. The advantage of JADE with exploitation and exploration by extreme individuals is shown by solving thirteen benchmark problems. 
研究論文(国際会議プロシーディングス)  共著  An Adaptive Differential Evolution with Learning Parameters According to Groups Defined by the Rank of Objective Values  Tetsuyuki Takahama and Setsuko Sakai  Proc. of the Eighth International Conference on Swarm Intelligence (ICSI2017)  411-419  2017/07  Differential Evolution (DE) has been successfully applied to various optimization problems. The performance of DE is affected by algorithm parameters such as a scaling factor F and a crossover rate CR. Many studies have been done to control the parameters adaptively. One of the most successful studies on controlling the parameters is JADE. In JADE, the values of each parameter are generated according to one probability density function (PDF) which is learned by the values in success cases where the child is better than the parent. However, search performance might be improved by learning multiple PDFs for each parameter based on some characteristics of search points. In this study, search points are divided into plural groups according to the rank of their objective values and the PDFs are learned by parameter values in success cases for each group. The advantage of JADE with the group-based learning is shown by solving thirteen benchmark problems. 
詳細表示...

研究発表
口頭発表(一般)  差分進化における相関係数に基づく遺伝子のグループ化とグループ単位の交叉の提案  情報処理学会 第120回数理モデル化と問題解決研究会  2018/09/26  最適化が困難な変数間依存性の強い問題に出現する特徴的な分布として,細い楕円形状の分布がある.このような場合に優れた子個体を生成するには,楕円形状の長軸に沿って変数を同時に変更する必要がある.また,同様の分布は,変数分離型の問題において探索点集合が最適解から離れている場合にも出現する.差分進化における2項交叉は,各変数(遺伝子) について同じ確率で交叉を行うかどうかを決定しているため,特定の遺伝子を同時に交叉することは困難である.本研究では,このような形状を検出するために探索点の相関係数を利用する方法を提案する.探索点の分布から相関行列を求め,相関の強い遺伝子をグループ化し,グループ単位で遺伝子を同時に交叉する(あるいは,交叉しない).本手法を差分進化の代表的手法であるJADE に導入し,幾つかのベンチマーク問題を最適化し,性能を比較することにより,本手法の効果を調べる. 
口頭発表(一般)  変数間依存性を解消する変換を導入したブレンド交叉の提案  京都大学数理解析研究所RIMS共同研究(公開型)「不確実性の下での意思決定理論とその応用 :計画数学の展開」  2017/11/15 
口頭発表(一般)  Particle Swarm Optimization with the Velocity Updating Rule According to an Oblique Coordinate System  The 2nd International Symposium on Swarm Behavior and Bio-Inspired Robotics (SWARM2017)  2017/10/30 
口頭発表(一般)  An Adaptive Differential Evolution with Exploitation and Exploration by Extreme Individuals  SICE Annual Conference 2017 (SICE2017)  2017/09/22 
口頭発表(一般)  An Adaptive Differential Evolution with Learning Parameters According to Groups Defined by the Rank of Objective Values  The Eighth International Conference on Swarm Intelligence (ICSI2017)  2017/07/27 
詳細表示...

受賞
IEEE Computational Intelligence Society, CEC2010 Constrained Real-Parameter Optimisation Competition Award  2010/07 
2006 IEEE Congress on Evolutionary Computation Competition Program Award -- Nonlinear Programming  2006/07 
詳細表示...

担当授業科目
機械学習特論 
データ構造とアルゴリズムⅠ 
数理計画法 
情報科学序説 
機械学習 
詳細表示...

所属学協会
電子情報通信学会 
言語処理学会 
人工知能学会 
情報処理学会 
IEEE  1999/05/01-現在 
詳細表示...

公開講座
模擬授業「自然から学ぶアルゴリズム-ナチュラル・コンピューティング-」  その他  2015/07-2015/07  自然界には,多くの有益なアイデアが存在しています.例えば,生物は長い年月をかけて進化し,環境に適応しています.また,単純な生物でも群としては驚くほど知的な振る舞いをします.このような現象をコンピュータ上に実現し,情報処理に応用する研究であるナチュラルコンピューティングを紹介します. 
模擬授業「自然から学ぶアルゴリズム-ナチュラル・コンピューティング-」  その他  2009/11-2009/11  自然界には,多くの有益なアイデアが存在しています.例えば,生物は長い年月をかけて進化し,環境に適応しています.また,単純な生物でも群としては驚くほど知的な振る舞いをします.このような現象をコンピュータ上に実現し,情報処理に応用する研究であるナチュラルコンピューティングを紹介します. 
模擬授業「人工知能」  公開講座  2008/10-2008/10  人工知能は,知的活動が可能な知的機械(コンピュータ)の実現を目指す研究分野です.推論,探索などの基礎分野から音声・画像認識,自然言語理解などの応用分野まで,人間の知的活動の一部を実現するために様々な研究が行われています.本講義では,ゲーム木探索を取り上げ,オセロゲームや将棋のようなゲームにおいて,指し手をコンピュータで計算する方法について,基本的な考え方を紹介します. 
平成17年度広島市立大学情報科学部公開講座  公開講座  2005/11-2005/11  自然から学ぶアルゴリズム -ナチュラル・コンピューティング- 
詳細表示...