Discriminant Dynamic Mode Decomposition for Labeled Spatio-Temporal Data Collections

  title={Discriminant Dynamic Mode Decomposition for Labeled Spatio-Temporal Data Collections},
  author={Naoya Takeishi and Keisuke Fujii and Koh Takeuchi and Yoshinobu Kawahara},
Extracting coherent patterns is one of the standard approaches towards understanding spatiotemporal data. Dynamic mode decomposition (DMD) is a powerful tool for extracting coherent patterns, but the original DMD and most of its variants do not consider label information, which is often available as side information of spatio-temporal data. In this work, we propose a new method for extracting distinctive coherent patterns from labeled spatio-temporal data collections, such that they contribute… 
1 Citations
Discovering spatiotemporal patterns of COVID-19 pandemic in South Korea
This work develops a novel framework of compatible window-wise dynamic mode decomposition (CwDMD) for nonlinear infectious disease dynamics and reveals that the first wave was contained in the Daegu and Gyeongbuk areas, but it spread rapidly to the whole of South Korea after the second wave.


Supervised dynamic mode decomposition via multitask learning
Learning Low-Dimensional Temporal Representations
This work jointly learns the subspace and the associated alignments by optimizing an objective which favors easily-separable temporal structures, and shows that this objective is connected to the inference of alignments, thus allows an iterative solution.
Discriminative Dimensionality Reduction for Multi-Dimensional Sequences
A novel supervised dimensionality reduction approach for sequence data, called Linear Sequence Discriminant Analysis (LSDA), which learns a linear discriminative projection of the feature vectors in sequences to a lower-dimensional subspace by maximizing the separability of the sequence classes such that the entire sequences are holistically discriminated.
Discriminative Transformation for Multi-Dimensional Temporal Sequences
A method to transform features in sequences into a low-dimensional subspace such that different sequence classes are holistically separated and provides a new tractable and effective solution with theoretical proofs by constraints unfolding and pruning, convex relaxation, and within-class scatter compression.
Dynamic mode decomposition - data-driven modeling of complex systems
This first book to address the DMD algorithm presents a pedagogical and comprehensive approach to all aspects of DMD currently developed or under development, and blends theoretical development, example codes, and applications to showcase the theory and its many innovations and uses.
Variable Projection Methods for an Optimized Dynamic Mode Decomposition
A simple algorithm for computing an optimized version of the dynamic mode decomposition for data which may be collected at unevenly spaced sample times and finds that the resulting decomposition displays less bias in the presence of noise than standard DMD algorithms.
Data-driven spectral decomposition and forecasting of ergodic dynamical systems
  • D. Giannakis
  • Mathematics
    Applied and Computational Harmonic Analysis
  • 2019
On dynamic mode decomposition: Theory and applications
A theoretical framework in which dynamic mode decomposition is defined as the eigendecomposition of an approximating linear operator, which generalizes DMD to a larger class of datasets, including nonsequential time series, and shows that under certain conditions, DMD is equivalent to LIM.
Discovering dynamic patterns from infectious disease data using dynamic mode decomposition
Dynamic mode decomposition (DMD) is a recently developed method focused on discovering coherent spatial-temporal modes in high-dimensional data collected from complex systems with time dynamics that is poised to be an effective and efficient computational analysis tool for the study of infectious disease.