2006 Fiscal Year Final Research Report Summary
Interactive Evolutionary Computation with Generative Interaction
| Project/Area Number |
17300072
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Sensitivity informatics/Soft computing
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
IBA Hitoshi The University of Tokyo, Graduate School of Frontier Sciences, Professor (40302773)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
ISHIZUKA Mitsuru The University of Tokyo, Graduate School of Information Science and Technology, Professor (50114369)
|
|---|
| Project Period (FY) |
2005 – 2006
|
| Keywords | Genetic Programming / Genetic Algorithms / Evolutionary Computation / Humanoid Robotics / Automatic Composition / Emergent Design / Generative Interaction / Interactive Evolutionary Computation |
|---|
| Research Abstract |
Evolutionary Algorithms have been applied to variety of problems and are considered to be powerful ways for solving problems. In this research, we have established an framework of generative interaction based on the interactive evolutionary computation. More precisely, we have demonstrated the effectiveness of the proposed approach in the following ways:
1. In the field of computational finances, we have designed a model based on Inductive Genetic Programming (IGP) for Time Series prediction that was able to predict price trends with remarkable precision. We have used this technique to understand the empirical differences between the stock and foreign exchange markets.
2. We proposed a new IEC interface which adopts an active musical playback interface for a composition-aid system. In the new interface, the system motivates its users, reducing their burden without the use of any filtering techniques. We performed experiments where subjects use the new interface to validate it. The result
… More
s showed that the new interface is better to motivate users to compose with an IEC system than other musical creation interfaces.
3. We have tried to achieve a method of intuitive motion design that could be used by people who have no specialized knowledge of robotics in order to operate humanoid robots. In our method, the desired motion is designed using IEC as an intuitive method of motion design for humanoid robots and is aimed at people with no specialized knowledge of robotics and with no large-scale equipment. The design of stable motions was achieved with the usage of the desired ZMP (i. e., zero moment point).
4. The number of modules, structure and controller are main parameters for the uniform modular robots. Among them, the number of modules decides amount of resources, robustness and fault tolerance. IEC technique has been successfully applied to resolve these design problems. We covered dynamic alteration of the optimal shapes by the number of modules. Our experimental results have shown the hill-climbing method is not fit to this problem. Less
|
Research Products
(26 results)
