"Information-Representation Optimization in High-dimensional Neural Networks" Project
Tohoku Univ., RIEC, Nation-wide Cooperative Research Project (Leaders: A.Hirose, S.Sato)
December 11-12, 2017, RIEC, Tohoku University

Tohoku Univ., RIEC, Nation-wide Cooperative Research Projects
Leaders: Akira Hiorse (The University of Tokyo), Shigeo Sato (Tohoku University)
December 11-12, 2017, RIEC, Tohoku University

December 11 (Mon), 2017
13:15 - 13:40 Hyunsoo Kim, Akira Hirose, "Polarimetric synthetic-aperture-radar land classification based on quaternion auto-encoder and quaternion self-organizing map"
13:40 - 14:05 Shigeo Sato, Yoshihiro Osakabe, "Quantum information processing and high-dimensional neural networks"
14:05 - 14:30 Eckhard Hitzer, Robert B. Easter, "Developing Conformal Geometric Algebra for the Representation of Algebraic Curves and Surfaces"
14:30 - 14:55 Masaki Kobayashi, "Singular points in complex-valued neural networks"
14:55 - 15:10 (coffee break)
15:10 - 15:35 Takehiko Mizoguchi, Isao Yamada, "A real domain computation for hypercomplex singular value decompositions and its application to tensor recovery"
15:35 - 16:00 Hiroki Kuroda, Masao Yamagishi, Isao Yamada, Teijiro Isokawa, Nobuyuki Matsui, "Exploiting Sparsity for Recovery of Finite-Dimensional Piecewise Continuous Signals"
16:00 - 16:25 Tohru Nitta, Yasuaki Kuroe, "Learning characteristics in hierachical hyperbolic neural networks"
16:25 - 16:50 Yasuaki Kuroe, "Recurrent spiking neural networks: Learning methods and their applications"

December 12 (Tue), 2017
09:30 - 12:00 Free discussion: High-dimensional neural networks -- Project Future Plan