The modular network Self-Organizing Map (mnSOM) we developed is a revolutionary information-processing algorithm closely resembling the structure of the cerebral cortex. An important advantage is its ability to map the processes executed by each module and produce an aerial perspective. The modules work well in a wide range of application fields because they can be freely exchanged to suit the application. In addition, because their architecture suits parallel computing, they can be used to build large-scale systems. We are currently conducting a broad range of applied research, including the realization of cooperation among autonomous robots, the development of calculation theory and new algorithms, and applications to hardware and parallel computing. This program covers the theory and applications of mnSOM, the core technology of our laboratory. Students have direct access to a wide range of valuable knowledge that cannot be found anywhere else. During the first half of the three-month program, students study classical SOM theory and programs while during the latter half, they study mnSOM programming and how to verify operation. We hope our students will use what they learn here in their own research and acquire the ability to check the quality of their results at a truly practical level. Having instructed students from many different labs, I have developed the contents of this program to thoroughly satisfy their needs. Please feel free to visit our laboratory anytime. To Page Top