Abstract

Everything that exists in the natural world is made up of several types of elementary particles. However, we cannot discover everything about nature by identifying the properties of these particles. This is because observations of nature reveal that the collection of many particles can give rise to wholly unpredictable phenomena, regardless of what we know about the particles’ properties. In the physical sciences, the importance of focusing on “a large number” of things is reflected in the expression, “More is different.” By introducing this perspective into information science, we promote research under the common concept of “More is different in informatics as well.” Preliminary studies have produced significant results regarding the basic theories of information and communications. In this project, we expect to deepen our achievements in these fields, while expanding into the fields of quantum information science and bioinformatics.

Research Categories

A) Information and communication theories

We will continue our cutting-edge research in issues in information and communication theories, where the methods of the statistical mechanics of information processing have yielded many new approaches to low-density parity check codes, code division multiple access systems, probabilistic image processing and other fields.

B) Quantum information

We will introduce methods of approximation from the statistical mechanics of information processing to quantum computation, a field which currently is developing at a tremendous rate, and will construct a many-body theory to employ in creating methods to use and control quantum probability.

C) Bioinformatics

We will introduce methods of approximation from the statistical mechanics of information processing for the analysis of bioinformatics, one of the most important uses of probabilistic models. We will establish a theoretical framework for the analysis of data from the view point of systems science, which will provide some guidance in creating models for the objects of our research.

D) Statistical mechanics of information processing

We will develop new statistical mechanics-based algorithms and analytical methods for probabilistic models. We will continue our investigations, incorporating approaches based on information geometry, algebraic geometry, and theoretical computer science, among other fields, and use them as a springboard for advances in informatics.