Course title
知能機械デザイン学特論   [Intelligent Machine Design]
Course category	courses for master's programs	Requirement		Credit	2
Department		Year	～	Semester	3rd
Course type	3rd	Course code	1060607
Instructor(s)
近藤 敏之   [KONDO Toshiyuki]
Facility affiliation	Faculty of Engineering	Office	afjgxte/L1151	Email address

Course description

- The purpose of the course is to understand computational intelligence methodology for designing intelligent machines that have adaptability, such as artificial neural networks, evolutionary computation, reinforcement learning, etc.

Expected Learning

The standard of this lecture is to understand computational intelligence methodologies such as artificial neural networks, evolutionary computation, reinforcement learning. See the Curriculum maps.

Course schedule

1. Introduction - Good Old Fashioned Artificial Intelligence (GOFAI) and Artificial Life 2. Neural Computation (1) - Adaptive Mechanism of Primitive Creature - Perceptron 3. Neural Computation (2) - Supervised learning of Neural Networks (Back Propagation method) 4. Neural Computation (3) - Hopfield networks model and Boltzmann Machine 5. Evolutionary Computation (1) - Introduction of Genetic Algorithms (GA) 6. Evolutionary Computation (2) - Real-coded GA and Multi-objective GA 7. Evolutionary Computation (3) - Other Evolutionary Computational Approach: EA, EP, GP, PSO, ACO 8. Reinforcement Learning (1) - Principle of Optimality, Markov Decision Process, State Value function 9. Reinforcement Learning (2) - Action Value function, Q-Learning, TD-Learning, Actor-Critic method 10. Reinforcement Learning (3) - Recent topics in RL 11. Unsupervised Learning (1) - Clustering 12. Unsupervised Learning (2) - Deep Learning 13. Design principle for Intelligent Machine (1) - Cognitive Psychology 14. Design principle for Intelligent Machine (2) - Brain Science, Huma Motor Learning 15. Design principle for Intelligent Machine (3) - BMI/BCI, Human-Robot interaction

Prerequisites

In addition to 30 hours that students spend in the class, students are recommended to prepare for and revise the lectures, spending the standard amount of time as specified by the University.

Required Text(s) and Materials

Digital textbooks are distributed in the class.

References

Assessment/Grading

Based on the score of assignments (programming and/or survey of research papers)

Message from instructor(s)

I believe that computational Intelligence that learns from living systems would be quite an interesting topic for the students in Computer Science.

Course keywords

Machine Learning, Neural Networks, Evolutionary Computation, Reinforcement Learning

Office hours

Ask questions at any time by email.

Remarks 1

Remarks 2

Related URL

This course uses Moodle

Lecture Language

Japanese

Language Subject

Last update

3/8/2019 9:51:47 PM