理工学研究科Graduate School of Science and Engineering
HUI500X3(人間情報学 / Human informatics 500)知的情報処理特論2Intelligent information processing (Ⅱ)
彌冨 仁Hitoshi IYATOMI
授業コードなどClass code etc
学部・研究科Faculty/Graduate school | 理工学研究科Graduate School of Science and Engineering |
添付ファイル名Attached documents | |
年度Year | 2022 |
授業コードClass code | YB017 |
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 | 応用情報工学専攻 |
すべて開くShow all
すべて閉じるHide All
Outline (in English)
This course emphasizes on important techniques and practical application of recent machine learning techniques including image understanding and natural language processing.
Then, the course introduces deep models based on generative models and explains their fundamentals.
Finally, this course pursuits several cutting-edge techniques associate with machine learning.
授業で使用する言語Default language used in class
英語 / English
授業の概要と目的(何を学ぶか)Outline and objectives
This course emphasizes on important techniques and practical application of recent machine learning techniques including image understanding and natural language processing.
Then, the course introduces deep models based on generative models and explains their fundamentals.
Finally, this course pursuits several cutting-edge techniques associate with machine learning.
到達目標Goal
Develop a comprehensive abilities for understanding state-of-the-art papers.
Develop an ability to find practical problems associated with machine learning fields (i.e. classification, regression, inference), find a path to their solutions and solve them.
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?
ディプロマポリシーのうち、「DP1」「DP2」「DP3」に関連
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. )
Each class usually consists of lecture, discussion and exercise. Students are requested to do exercise in each class and some homework assignments. Classes in the latter part of this course are dedicated for presentation and discussion on cutting-edge papers.
# Contents may vary according to proficiency of students.
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。
1[オンライン/online]:Introduction of generative models
Introduction of generative adversarial networks (GANs) and related techniques
2[対面/face to face]:Fundamental of generative models (1)
[with exercise] Bayesian probabilistic theory, ML, MAP estimation,
conjugate prior,
3[対面/face to face]:Fundamental of generative models (2)
[with exercise] Convex set and convex function, Jensen’s inequality
4[対面/face to face]:Fundamental of generative models (3)
[with exercise] Expectation maximization (EM) algorithm, Kullback-Leibler (KL) divergence, [with exercise]
5[対面/face to face]:Fundamental of generative models (4)
[with exercise] EM algorithm – Bayes estimation
6[対面/face to face]:Dimensional reduction and generative model
Manifold hypothesis, Variational autoencoders (VAE)
7[対面/face to face]:Generative adversarial networks (GAN) (1)
Introduction of GAN
8[対面/face to face]:Generative adversarial networks (GAN) (2)
Deep convolutional GAN, Wasserstein GAN, evaluation criteria (inception score, Fréchet inception distance (FID)), introduction of various applications
9[対面/face to face]:Semi-supervised training (1)
Introduction, classical method (self training, co-training, graph-based training), siamese-, triplet- networks. noisy students
10[対面/face to face]:Semi-supervised training (2)
Adversarial example, adversarial training and virtual adversarial training (VAT)
11[対面/face to face]:Transformers
Attention mechanism and self-attention mechanism – Important base model for various applications
12[対面/face to face]:BERT
Introduction of BERT – Why is it so hailed?
13[対面/face to face]:Vision Transformers
Introduction of vision transformers - It could be a model to replace CNN.
14[対面/face to face]:summary
sumamry
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.)
【本授業の準備・復習時間は、各4時間を標準とします。】Students must have basic background of machine learning.
テキスト(教科書)Textbooks
No specific textbook assigned. Materials (paper, reference, book chapter, slides) will be provided from time to time.
Students will find good references by their own.
参考書References
"Pattern recognition and machine learning" C.Bishop, Springer 2006.
Keras documentation (deep learning framework)
https://keras.io/
成績評価の方法と基準Grading criteria
60% in exercises in class and homework
40% in final report
(both on-line, off-line)
Use Hoppii system (Learning support system) to submit assignments and feedback.
Explanations of the assignments (except for the final assignment) will be given in the next and subsequent classes.
学生の意見等からの気づきChanges following student comments
Support with native language is sometimes necessary.
学生が準備すべき機器他Equipment student needs to prepare
Own computer (use google colaboratory)
その他の重要事項Others
Use google colaboratory.
Minimum programming skills in Python is required.
If the course is offered online, changes in online class methods, plans, and grading methods will be presented on a case-by-case basis in the Learning support system (Hoppii). Please check carefully on a daily basis to see if your instructor contacts you via the learning support system.