index＞What is U-Mart?
What is U-Mart?
As one of Japan's top artificial market research projects, many researchers join this project and do various activities. The major objective of this project is to design a financial market system. More specifically, we hope to establish market control methods by controlling the extent and scope of information disclosure with elements such as circuit breakers including commission rate and price movement limits, market maker, indicative price calculation methods, and changes in information update intervals. We measure information value and information tradeoffs (e.g. liquidity or stability) to do basic research to use information disclosure timing and scope as market control parameters.
U-Mart System is used as an excellent courseware for engineering and economics. Engineering educational institutions use the U-Mart system for programming practices. Also the investment program is very useful for practice because it can actually be operated, is open to set objectives from very simple algorithms to sophisticated learning algorithms, and motivates students by providing opportunities for competitions and other events (open experiments).
Experiments open to the public inviting public machine agents and human agents, and related discussions inviting experts from various fields are categorized as event activities. In recent years, UMIE 20xx series international open experiments and U-Mart20xx series domestic open experiments are periodically held. Also, we have special or tutorial sessions at international and domestic conferences hosted by NAACSOS, ISAGA, Japan Association for Evolutional Economics, and Information Processing Society of Japan. At each event, we collect and report the results of U-Mart research as well as provide experts from various fields with discussion opportunities by having panel discussions.
These three types of activities are inseparably tied. Part of their relationships is shown in the above figure. Machine agents invited to open experiments are necessary to increase the diversity of agent sets used for research. The diversity motivates researchers from various fields to have symposiums and workshops to gather and it serves as a springboard for another open experiment or joint research. The set of tools developed for educational purposes are used for research and events. Many machine agents have been developed through educational courses and have contributed to agent sets used for further research. More economics students join U-Mart as human agents, and more experiment opportunities are given. Also, the students have proposed GUI improvement ideas and have contributed to help develop new tools used for event activities. As more open experiments are held, more problems must be resolved using the artificial market are found as well as logs to be analyzed. Also as the research progressed more, purposes of open experiments have become clearer, and rules & systems have changed.