|
|
|
|
|
|
|
|
|
One of the most interesting features of intelligent systems is that it lies on the boundary of several academic disciplines, computer sciences, statistics, mathematics, and engineering. Over the past ten years, this inherently multi-disciplinary, from finance to biology and medicine to physics and chemistry and beyond, has been embraced and understood, with many benefits for researchers in the field.
Intelligent systems are usually studied as part of artificial intelligence (AI), which puts it firmly in computer science, and given the focus on the algorithm it certainly fits there. Especially, machine learning (ML), which is about making computer modify or adapt their actions (where these actions are making predictions, or controlling a robot) can be applied varies even more widely area.
The objective of this class is to lecture you modern AI approach with a central focus on several useful ML algorithms and theories and related matters. You can learn about the basic techniques and tricks of ML, and would apply them to your study or work.
|
|
|
1. Introduction: What is the machine learning?
2. Supervised learning(1): linear discriminant.
3. Supervised learning(2): Perceptron and neural networks.
4. Supervised learning(3): Radial basis functions.
5. Supervised learning(4): Support vector machine.
6. Practical application case study (1)
7. Unsupervised learning(1): k-means clustering, vector quantization.
8. Unsupervised learning(2): Self-organization map and related works.
9. Practical application case study (2)
10. Reinforcement learning (1): Dynamic programming and Temporal differences.
11. Reinforcement learning (2): Exploration and exploitation algorithm.
12. Practical application case study (3)
13. Evolutionary learning: Genetic algorithm.
14. Markov chain Monte Carlo methods.
15. Hot issues of machine learning.
|
|
|
|
Impremenation skill of python and R
|
|
|
|
To enable design and implementation a machine learning application for real-world problems.
|
|
|
|
To enable explain a machine learning algorithms for typical problem class.
|
|
|
1. Quiz score 10×score(0〜4) = max 40%
2. Reporting assignment ( 3 reports) 3×score(0〜20) = max 60%
|
|
|
【成績評価の基準表】
| 秀(S) | 優(A) | 良(B) | 可(C) | 不可(F) |
| 履修目標を越えたレベルを達成している | 履修目標を達成している | 履修目標と到達目標の間にあるレベルを達成している | 到達目標を達成している | 到達目標を達成できていない |
|
|
履修目標：授業で扱う内容（授業のねらい）を示す目標
到達目標：授業において最低限学生が身につける内容を示す目標
|
|
|
|
|
|
|
|
|
|
Machine learning, Artificial intelligence
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|