Center for Advanced Intelligence Project

[CLOSED]
We pursue the methodology of statistics and machine learning. Statistics has been playing important roles as a theoretical basis for data science and artificial intelligence. It provides the methodology of inductive inference by considering probability. We believe that working on real data analysis will lead to the development of theory and methods of statistics. We developed a method of statistical hypothesis testing (multiscale bootstrap) which is now commonly used for DNA sequence analysis and gene expression analysis. We also developed a theory of information criterion for the transfer learning (covariate shift) of machine learning. Recently, we are also working on statistical inference of the growth mechanism of complex networks, and multivariate analysis methods and their deep learning for integrating several types of data such as images and sentences.