*See the 2021 syllabus for the class held on 2021.
Living systems adapts their environment flexibly. Functions of information processing in the living system have been modeled in biomimetic computational systems. The biomimetic computational systems effectively solve problems which are difficult to be modeled mathematically or have nonlinearity, with which classical computational methods has difficulty. Some of the biomimetic computational systems have been established in engineering. In this course, students learn elementary techniques of artificial neural network, which is gaining attention as artificial intelligence and deep learning in recent years, reinforcement learning, genetic algorithms, their mathematical foundation and their application to practical problems. Additionally, the students learn analyses for experimental data of biological information.
Living systems adapts their environment flexibly. Functions of information processing in the living system have been modeled in biomimetic computational systems. The biomimetic computational systems effectively solve problems which are difficult to be modeled mathematically or have nonlinearity, with which classical computational methods has difficulty. Some of the biomimetic computational systems have been established in engineering. In this course, students learn elementary techniques of artificial neural network, which is gaining attention as artificial intelligence and deep learning in recent years, reinforcement learning, genetic algorithms, their mathematical foundation and their application to practical problems. Additionally, the students learn analyses for experimental data of biological information.
The purpose of this course is to learn biomimetic computational systems such as artificial neural networks, reinforcement
learning, genetic algorithms, their mathematical foundation, their application to practical problems, and analyses for experimental
data of biological information.
- Can understand mathematical foundation of biomimetic computational systems
- Can apply biomimetic computational systems to practical problems
- Can understand analyses for experimental data of biological information
| Class schedule | HW assignments (Including preparation and review of the class.) | Amount of Time Required | |
|---|---|---|---|
| 1. | Guidance - Biomimetic computational systems Basics of Machine learning 1 - Supervised learning |
assignments distributed in the class | 190minutes |
| 2. | Basics of Machine learning 2 - Regression analysis - Discriminant analysis |
assignments distributed in the class | 190minutes |
| 3. | Basics of Machine learning 3 - Optimization techniques |
assignments distributed in the class | 190minutes |
| 4. | Neural information processing 1 - Formal neuron - Perceptron |
assignments distributed in the class | 190minutes |
| 5. | Neural information processing 2 - Hierarchical neural networks - Backpropagation algorithm |
assignments distributed in the class | 190minutes |
| 6. | Neural information processing 3 - Time series data - Recurrent neural networks |
assignments distributed in the class | 190minutes |
| 7. | Neural information processing 4 - dynamical systems and nonlinear dynamics - Hopfield type neural networks |
assignments distributed in the class | 190minutes |
| 8. | Neural information processing 5 - Chaos dynamical systems - Chaotic neural networks |
assignments distributed in the class | 190minutes |
| 9. | Neural information processing 6 - Boltzmann Machine |
assignments distributed in the class | 190minutes |
| 10. | Reinforcement learning - Basics of reinforcement learning - Application of reinforcement learning |
assignments distributed in the class | 190minutes |
| 11. | Genetics and information processing - Genetic algorithm - Mathematical models of population genetics |
assignments distributed in the class | 190minutes |
| 12. | Analyses for experimental data of biological information 1 - Statistical evaluation |
assignments distributed in the class | 190minutes |
| 13. | Analyses for experimental data of biological information 2 - Analysis of data in brain-computer interface, as an example - Application of biomimetic computational systems |
assignments distributed in the class | 190minutes |
| 14. | Examination and review -Examination coverage: all topics |
Summarize all topics | 190minutes |
| Total. | - | - | 2660minutes |
| four assignments | examination | Total. | |
|---|---|---|---|
| 1. | 25% | 15% | 40% |
| 2. | 25% | 15% | 40% |
| 3. | 10% | 10% | 20% |
| Total. | 60% | 40% | - |
-Grade is judged by four assignments (60%) and the result of examination (40%).
The level of 60 points is to understand basics of mathematical foundation of biomimetic computational systems, their application to practical problems of and analyses for experimental data of biological information.
The level of 60 points is to understand basics of mathematical foundation of biomimetic computational systems, their application to practical problems of and analyses for experimental data of biological information.
Presentation materials in each class will be uploaded online.
Reference books (in Japanese)
瀧 雅人，”機械学習スタートアップシリーズ これならわかる深層学習入門 (KS情報科学専門書)”，講談社
Reference books (in Japanese)
瀧 雅人，”機械学習スタートアップシリーズ これならわかる深層学習入門 (KS情報科学専門書)”，講談社
- Lunch break on Friday: Making appointment is recommended.
- Course that cultivates an ability for utilizing knowledge
Last modified : Sun Mar 21 14:53:25 JST 2021