Reproducing kernel Hilbert C*-module and kernel mean embeddings
Preprint
Abstract
Kernel methods have been among the most popular techniques in machine learning, where learning tasks are solved using the property of reproducing kernel Hilbert space (RKHS). In this paper, we propose a novel data analysis framework with reproducing kernel Hilbert $C^*$-module (RKHM) and kernel mean embedding (KME) in RKHM. Since RKHM contains richer information than RKHS or vector-valued RKHS (vvRKHS), analysis with RKHM enables us to capture and extract structural properties in such as functional data. We show a branch of theories for RKHM to apply to data analysis, including the representer theorem, and the injectivity and universality of the proposed KME. We also show RKHM generalizes RKHS and vvRKHS. Then, we provide concrete procedures for employing RKHM and the proposed KME to data analysis.Keywords
All Related Versions
This publication has 19 references indexed in Scilit:
- Koopman Operator Spectrum for Random Dynamical SystemsJournal of Nonlinear Science, 2019
- Dynamic mode decomposition in vector-valued reproducing kernel Hilbert spaces for extracting dynamical structure among observablesNeural Networks, 2019
- Eigendecompositions of Transfer Operators in Reproducing Kernel Hilbert SpacesJournal of Nonlinear Science, 2019
- Deep learning for universal linear embeddings of nonlinear dynamicsNature Communications, 2018
- Theory of Reproducing Kernels and Applications2016
- Operator-valued kernel-based vector autoregressive models for network inferenceMachine Learning, 2014
- Local convergence of the steepest descent method in Hilbert spacesJournal of Mathematical Analysis and Applications, 2004
- HilbertC *- andW *-modules and their morphismsJournal of Mathematical Sciences, 2000
- C*-Algebras and Hilbert Space Operators1990
- Probabilistic and Statistical Aspects of Quantum TheoryPhysics Today, 1984