Systems, in which a huge number of simple elements cooperate with each other with randomness, could have unexpected capacities. This is similar to unpredictable phenomena in natural world that are created by a lot of simple molecules with interactions and fluctuations. Such mechanism is often called "more is different" in physics. One of the biggest features of the probabilistic information processing is to capture the mechanism of "more is different" in natural world by physical and mathematical viewpoints, and to apply the mechanism and mathematical techniques that have been sophisticated in the physical long history to computer science. In our laboratory, we promote researches under the common concept of "more is different in informatics as well."

  Our goal is to establish strong theoretical paradigms for the probabilistic information processing to affect the artificial intelligence, Big Data science, the pattern recognition and life information science as well as the signal processing by means of statistical science, statistical physics and quantum mechanics.