Masaki Kawamura Prof.


Masaki Kawamura

Graduate School of Sciences and Technology for Innovation, Yamaguchi University

College of Science

Department of Informatics

E-mail m.kawamu[AT]
Web Sites
Specializations Information Processing Algorithm by Statistical Mechanics
Research Interests Information processing models, e.g., neural network models and information hiding models, are investigated on statistical mechanics. We evaluate the memory capacity in an associative memory model, which is a type of neural network model, and the channel capacity in a digital watermarking model. These quantitative evaluations are required for the theoretical approach. Although applications of these models seem to be different, they have similar mathematical structure or Hamiltonian. Therefore, these models can be analyzed by the same statistical mechanics theoretically. Moreover, the optimal algorithms can be derived from the Hamiltonian for novel information processing models.