In addition to the machine learning methods learned in "EEE.M251: Electrical and Electronic Informatics I", this course (EEE.M252) covers various algorithms for deep learning. An understanding of the basic concepts of deep learning is important for future research and development in various fields of electrical and electronic systems.
Students will acquire the following abilities.
1) Deepen understanding of various algorithms for deep learning.
2) To be able to code a simple neural network in Python.
AI, deep learning, Python
✔ Specialist skills | Intercultural skills | Communication skills | ✔ Critical thinking skills | Practical and/or problem-solving skills |
✔ Applied professional capacity in electrical and electronic fields |
In each lecture, explanations will be given using specified slides.
In addition, there will be 2-3 guest lectures.
Course schedule | Required learning | |
---|---|---|
Class 1 | Introduction | Introduction |
Class 2 | Concepts of Deep Learning | Deepen understanding of the overall overview of deep learn |
Class 3 | Loss function | Learn about typical loss function used for deep networks |
Class 4 | Back Propagation method | Learn about Back Propagation method. |
Class 5 | Convolutional neural networks (CNN) | Learn the basics of convolutional neural networks for image data processing |
Class 6 | Vanishing gradient problem and Convergence in deep learning I | Learn about vanishing gradient problem and convergence. |
Class 7 | Vanishing gradient problem and Convergence in deep learning II | Learn about vanishing gradient problem and convergence. |
Class 8 | Guest Lecture | Guest Lecture |
Class 9 | Recurrent neural network | Learn about recursive neural networks for time series data processing |
Class 10 | Attention mechanism and Transformer I | Explain the details of the Attention Mechanism and Transformer. |
Class 11 | Attention mechanism and Transformer II | Explain the details of the Attention Mechanism and Transformer. |
Class 12 | Guest Lecture | Guest Lecture |
Class 13 | Generative model I | Learn about variational auto-encoder (VAE). |
Class 14 | Generative model II | Learn about typical diffusion model. |
Class 15 | Guest Lecture | Guest Lecture |
Basically, no preparation is required, but it is strongly recommended to read through the relevant sections of reference books to deepen your understanding.
In each lecture, explanations will be given using specified slides.
Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems (O'Reilly)
深層学習 改訂第2版 (機械学習プロフェッショナルシリーズ)(講談社)(in Japanese)
拡散モデル データ生成技術の数理(岩波書店)(in Japanese)
Grades will be assigned as follows.
Attendance report (40%)
Midterm: Submission of Python code (30%)
Final exam (30%)
Just a little motivation and passion
EEE.M251 Electrical and Electronic Informatics I : Near required
EEE.M221 Computation Algorithms and Programming : Recommended
EEE.M231 Applied Probability and Statistical Theory : Recommended