Dimensionality reduction of collective motion by principal manifolds
From MaRDI portal
Publication:282231
DOI10.1016/J.PHYSD.2014.09.009zbMATH Open1360.94130arXiv1508.03332OpenAlexW1967014337MaRDI QIDQ282231FDOQ282231
Authors: Kelum Gajamannage, Sachit Butail, M. Porfiri, Erik Bollt 
Publication date: 12 May 2016
Published in: Physica D (Search for Journal in Brave)
Abstract: While the existence of low-dimensional embedding manifolds has been shown in patterns of collective motion, the current battery of nonlinear dimensionality reduction methods are not amenable to the analysis of such manifolds. This is mainly due to the necessary spectral decomposition step, which limits control over the mapping from the original high-dimensional space to the embedding space. Here, we propose an alternative approach that demands a two-dimensional embedding which topologically summarizes the high-dimensional data. In this sense, our approach is closely related to the construction of one-dimensional principal curves that minimize orthogonal error to data points subject to smoothness constraints. Specifically, we construct a two-dimensional principal manifold directly in the high-dimensional space using cubic smoothing splines, and define the embedding coordinates in terms of geodesic distances. Thus, the mapping from the high-dimensional data to the manifold is defined in terms of local coordinates. Through representative examples, we show that compared to existing nonlinear dimensionality reduction methods, the principal manifold retains the original structure even in noisy and sparse datasets. The principal manifold finding algorithm is applied to configurations obtained from a dynamical system of multiple agents simulating a complex maneuver called predator mobbing, and the resulting two-dimensional embedding is compared with that of a well-established nonlinear dimensionality reduction method.
Full work available at URL: https://arxiv.org/abs/1508.03332
Recommendations
- Parsimonious representation of nonlinear dynamical systems through manifold learning: a chemotaxis case study
- Principal Manifolds and Nonlinear Dimensionality Reduction via Tangent Space Alignment
- Nonlinear Dimension Reduction by Local Multidimensional Scaling
- Nonlinear dimensionality reduction by topologically constrained isometric embedding
- A metric multidimensional scaling-based nonlinear manifold learning approach for unsupervised data reduction
Cites Work
- Principal component analysis.
- Diffusion maps
- A note on two problems in connexion with graphs
- Introduction to algorithms
- Principal Curves
- Title not available (Why is that?)
- An Algorithm for Finding Best Matches in Logarithmic Expected Time
- Hessian eigenmaps: Locally linear embedding techniques for high-dimensional data
- A textbook of graph theory
- Analysis and classification of collective behavior using generative modeling and nonlinear manifold learning
- ATTRACTOR MODELING AND EMPIRICAL NONLINEAR MODEL REDUCTION OF DISSIPATIVE DYNAMICAL SYSTEMS
- Title not available (Why is that?)
- Portraits of self-organization in fish schools interacting with robots
- Parameter Selection for Principal Curves
- Advances in Neural Networks – ISNN 2005
Cited In (5)
- Recurrent neural networks for dynamical systems: applications to ordinary differential equations, collective motion, and hydrological modeling
- Parsimonious representation of nonlinear dynamical systems through manifold learning: a chemotaxis case study
- Detecting phase transitions in collective behavior using manifold's curvature
- Analysis and classification of collective behavior using generative modeling and nonlinear manifold learning
- Collective pulsing in xeniid corals. I: Using computer vision and information theory to search for coordination
Uses Software
This page was built for publication: Dimensionality reduction of collective motion by principal manifolds
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q282231)
