講義科目名		Computational Intelligence
科目ナンバリングコード		DEG-HSI7311E
講義題目
授業科目区分		大学院科目　Subjects for Graduate School of Design
開講年度		2021
開講学期		春学期
曜日時限		春学期 月曜日 １時限 春学期 月曜日 ２時限
必修選択		選択　Elective
単位数		2
担当教員		高木　英行
開講学部・学府		芸術工学府
対象学部等		芸術工学府　デザイン人間科学国際コース　Department of Design，Human Science International Course
対象学年		博士2年/博士3年　Doctor second grade/third grade
開講地区		大橋地区
その他 （自由記述欄）

履修条件		Eligibility: This is a lecture course for doctoral students who enrolled in April, 2020 and before.
授業概要		We learn three computational intelligence technologies including evolutionary computation, fuzzy systems, and artificial neural networks, mainly and other techniques including knowledge engineering and artificial life using the below textbook and materials delivered at class.
授業形態 （項目）		☑　lecture・exercise □　lab works □　group works ・ pair works □　field exercise ☑　presentations □　discussions
授業形態 （内容）		No.1 - No.12 classes: lectures No.13 - No.15 classes: presentations; see the below lecture plans.
使用する教材等		writing on a blackboard, textbook, slides, and other media
全体の教育目標		The objective of this course is to let students obtain the applicability of these technologies for their Master/Doctoral research.
個別の教育目標		Students should come to be able to: (1) explain the optimization mechanism of evolutionary computation, (2) calculate genetic operations on paper, (3) calculate fuzzy reasoning of fuzzy systems that has more than equal two inputs, and (4) explain learning mechanism of neural network with both supervised leaning and unsupervised learning.
授業計画		1. Artificial Life and Complex Systems * We understand both artificial life and complex systems as the systems that generate complex outputs based on simple mechanism. This means that we may be able to understand a simple fundamental from complex observations. We learn cellular automaton, life game, L-system, and BOID. 2-4 Evolutionary Computation * Introduction to evolutionary computation. * Practice of Genetic Algorithms (GA) operations. * Other techniques including Genetic Programming, Niche-GA, Differential Evolution, Evolutionary multi-objective Optimization. 5. Design based on Human KANSEI * Interactive Evolutionary Computation (IEC) is an evolutionary computation that optimizes a target system based on human user's subjective evaluation for the output from the target system. We glance many IEC applications and lean its wide applicability. 6-8 Fuzzy Systems * We glance how fuzzy systems have been used in real world. * We learn how to realize fuzzy reasoning that we human being can using fuzzy information/data in computer. It includes fuzzy sets, fuzzy reasoning rules, fuzzy reasoning methods. * Practice of fuzzy reasoning 9-11 Neural Networks * Introduction to neural networks (NN): concept, biological models, NN calculation, mechanism of NN's nonlinearity, and NN applications. * Supervised learning algorithms train NN using training data consisting of input data and supervised data as if teachers make students learning through teaching. We learn backpropagation algorithm and NN programming. * Unsupervised learning algorithms train NN without teaching as if we came to be able to speak and walk when we are children. We study competitive learning and Self-Organizing Map and how they cluster data without supervising. 12. Fusion Technologies * We learn cooperative models of evolutionary computation, fuzzy systems, and neural networks and their effectiveness that comes from combined strong points of each model. 13-15 Common View to Computational Intelligence Techniques and its Application to Master Research * We view evolutionary computation, fuzzy systems, and neural networks from the nonlinear point of view and understand the common feature. * We discuss how to apply studied computational intelligence techniques to Master/Doctoral Research of each student.
キーワード		evolutionary computation, genetic algorithms, fuzzy systems, neural networks
授業の進め方		This course consists of course lectures mainly and exercise time.
テキスト		Taguchi, Mitsui, and Takagi, "Keywords of Complexity", Kyoritsu Shuppan Co., Ltd. (2000) (in Japanese). For those who are hard to read Japanese, English materials are available; request them to Prof. Takagi.
参考書		For those who are hard to read Japanese books, Andries P. Engelbrecht , "Computational Intelligence: an introduction," Willey (2003 version) (2007 version)
学習相談		When you have questions, you may use Office Hour time or visit Prof. Takagi's office whenever he is there. Taking appointments by e-mails or phones is recommended.
試験／成績評価の方法等		final exam. or report (80%), ordinal attitude (10%), home assignments & small test (10%)
その他		Contact: Prof. TAKAGI, Hideyuki （takagi@design.kyushu-u.ac.jp）
添付ファイル
更新日付		2021-04-09 16:32:07.402