Professor of Information Science

YASUDA Muneki

 “Statistical machine learning theory,” the cornerstone of research

I am researching data science theory, based on statistical machine learning theory. Statistical machine learning theory is machine learning theory that has probability and statistical models as its theoretical pillars, and it occupies a key position in current AI and data science research.
In the first place, probability and statistics are tools that have been used by humans since ancient times to predict the unknown from known data, and to grasp the overall trends of a large amount of data. Statistical machine learning theory is a fusion of time-honored concepts from ancient times and modern computer science concepts.

Creating new value

We are tackling a wide range of data science problems, which includes AI, in the interdisciplinary areas of mathematics, probability and statistics, physics, and computer science. The main scope of our work is basic theoretical research in data science. In order to make more useful applications a reality, the basic flow of research is to solve mathematical equations and build algorithms, then demonstrate those results through programming. Since we have made developing the fundamental part of data science our main scope, it is a research field that can produce completely new forms of applications, which have heretofore never existed.

Theoretical research and applications

Due to the nature of our research field, the results of the research we conduct can be applied in many parts of the world. If I were to say that we are developing libraries that are helpful for developing more useful applications, people in the know may understand what I mean. Since the math, probability, and statistics that we learned in junior high school, high school, and university are actually serving a purpose out in the world in exactly the same forms we learned them, it is a very interesting area of research.