情報科学研究科Graduate School of Computer and Information Sciences
HUI500K1（人間情報学 / Human informatics 500）Advanced AIAdvanced AI
人工知能特論
黄 潤和Huang RUNHE
授業コードなどClass code etc
| 学部・研究科Faculty/Graduate school | 情報科学研究科Graduate School of Computer and Information Sciences |
| 添付ファイル名Attached documents | |
| 年度Year | 2024 |
| 授業コードClass code | TZ008 |
| 旧授業コードPrevious Class code | |
| 旧科目名Previous Class title | |
| 開講時期Term | 秋学期授業/Fall |
| 曜日・時限Day/Period | 水2/Wed.2 |
| 科目種別Class Type | |
| キャンパスCampus | 小金井 |
| 教室名称Classroom name | 各学部・研究科等の時間割等で確認 |
| 配当年次Grade | |
| 単位数Credit(s) | 2 |
| 備考（履修条件等）Notes | |
| 実務経験のある教員による授業科目Class taught by instructors with practical experience | |
| カテゴリーCategory |
すべて開くShow all
すべて閉じるHide All
授業の概要と目的（何を学ぶか）Outline and objectives
The topics include
・explaining advanced AI techniques
・introducing the state of the art AI techniques and researches
・solving some small problems in our daily life or for supporting the weak persons
到達目標Goal
The objectives of this course are to make students master the basic principles of AI, learn advanced AI techniques, know the state of the art AI researches, and able to solve a small problem with what they have learnt.
この授業を履修することで学部等のディプロマポリシーに示されたどの能力を習得することができるか（該当授業科目と学位授与方針に明示された学習成果との関連）Which item of the diploma policy will be obtained by taking this class?
Among diploma policies, "DP1" and "DP2" are related.
授業で使用する言語Default language used in class
英語 / English
授業の進め方と方法Method(s)（学期の途中で変更になる場合には、別途提示します。 /If the Method(s) is changed, we will announce the details of any changes. ）
This course is conducted by reviewing the basic AI techniques, and then students are asked to design a simple intelligent agent system based on the PAGE design components for solving a simple problem with a selected AI techniques and make the mid-term presentation. Students will receive the advice and comments during their presentation, and critical points are discussed among the professor and all students. Further students will step on learning some the state of the art AI technologies and researches from the lectures, selected research readings, and related video clips for further understanding. Finally students are to conduct a term project by freely selecting an application such as developing a practical intelligent system or designing a prototype smart system for a particular small problem in the real world. Students are asked to submit a final report on their term and make a presentation.
Moreover, there will be some questions for students to think and discuss in-class for promoting active learning and mutual learning among students as well.
アクティブラーニング（グループディスカッション、ディベート等）の実施Active learning in class (Group discussion, Debate.etc.)
あり / Yes
フィールドワーク（学外での実習等）の実施Fieldwork in class
なし / No
授業計画Schedule
授業形態/methods of teaching：対面/face to face
※各回の授業形態は予定です。教員の指示に従ってください。
第1回[対面/face to face]：Introduction
Overview of AI, History of AI, and the state of the art of AI
第2回[対面/face to face]：Problem solving
Searching for solutions to a problem
第3回[対面/face to face]：Application of problem solving
Each student selects a search problem topic to solve with a problem solving algorithm
第4回[対面/face to face]：Logic Reasoning
Knowledge representation and inference mechanisms
-forward chaining
-backward chaining
第5回[対面/face to face]：A decision-making system
A rule based system
-identification decision tree
-from a training data to a decision tree
-from a decision tree to refined rules
第6回[対面/face to face]：Decision-making under uncertainty
Each student proposes and implements a small decision-making system with learnt reasoning mechanisms
第7回[対面/face to face]：Mid-term presentation
Students make presentation of their work on design of an autonomous driving agent system
第8回[対面/face to face]：Big data mining
clustering, classification, knowledge discovery
第9回[対面/face to face]：Machine learning
Supervised/unsupervised/semi-supervised learning and transfer learning
第10回[対面/face to face]：Neural Network and Deep Learning and other hot topics
Neuron Networks, Back Propagation, CNN, LSTM, Transformer, AGI
第11回[対面/face to face]：Introduction of OpenAI and ChatGPT
Discovery of what functions in OpenAI and ChatGPT
第12回[対面/face to face]：Decision making under uncertainty;
Tesla autonomous driving system
Explaining Bayesian Rule and Markova chaining Models; Understanding Tesla's full self-driving (FSD) system
第13回[対面/face to face]：Discussion on the term project
Students report their progress of the term project, group discussion, and are given advice and suggestions
第14回[対面/face to face]：Final term project presentation
Students makes presentation of their term projects
授業時間外の学習（準備学習・復習・宿題等）Work to be done outside of class (preparation, etc.)
Students should be able to search for related research articles and read them. They are asked to identify a small problem in our daily life to solve for the final term-project.
It takes four hours for weekly pre-study and assignments on average.
テキスト（教科書）Textbooks
[1]. “Artificial Intelligence – A Modern Approach”, Stuart Russell and Peter Norvig, Prentice Hall.
[2]. "Deep Learning", MIT press, Ian Goodfellow and Yoshua Bengio and Aaron Courville, https://www.deeplearningbook.org/
参考書References
[1]. Neural networks and deep learning related web sites
[2]. OpenAI and ChatGPT topic related web sites
[3]. LLM and Transformer Models
成績評価の方法と基準Grading criteria
Evaluation on students is based on the mid-term presentation and report(30%)and final term-project presentation and report(70%).
学生の意見等からの気づきChanges following student comments
Interactive learning is preferable.
Interactive discussion in classes