Elastic principal manifolds and their practical applications

Abstract

Principal manifolds serve as useful tool for many practical applications. These manifolds are defined as lines or surfaces passing through "the middle" of data distribution. We propose an algorithm for fast construction of grid approximations of principal manifolds with given topology. It is based on analogy of principal manifold and elastic membrane. The first advantage of this method is a form of the functional to be minimized which becomes quadratic at the step of the vertices position refinement. This makes the algorithm very effective, especially for parallel implementations. Another advantage is that the same algorithmic kernel is applied to construct principal manifolds of different dimensions and topologies. We demonstrate how flexibility of the approach allows numerous adaptive strategies like principal graph constructing, etc. The algorithm is implemented as a C++ package elmap and as a part of stand-alone data visualization tool VidaExpert, available on the web. We describe the approach and provide several examples of its application with speed performance characteristics.

26 pages, 10 figures, edited final version

References in corpus (3)

• Constructive Methods of Invariant Manifolds for Kinetic Problems
• Invariant grids for reaction kinetics
• Neural network modeling of data with gaps: method of principal curves, Carleman's formula, and other

Cited by in corpus (19)

• Principal Graphs and Manifolds
• Principal manifolds and graphs in practice: from molecular biology to dynamical systems
• Robust And Scalable Learning Of Complex Dataset Topologies Via Elpigraph
• T-ReX: a graph-based filament detection method
• Topological Grammars for Data Approximation
• Quasi-Equilibrium Closure Hierarchies for The Boltzmann Equation
• Order--disorder separation: Geometric revision
• Data complexity measured by principal graphs
• Personality Traits and Drug Consumption. A Story Told by Data
• ViDaExpert: user-friendly tool for nonlinear visualization and analysis of multidimensional vectorial data
• Elastic Maps and Nets for Approximating Principal Manifolds and Their Application to Microarray Data Visualization
• Piece-wise quadratic approximations of arbitrary error functions for fast and robust machine learning
• Learning from Successful and Failed Demonstrations via Optimization
• Coloring Panchromatic Nighttime Satellite Images: Comparing the Performance of Several Machine Learning Methods
• PCA Beyond The Concept of Manifolds: Principal Trees, Metro Maps, and Elastic Cubic Complexes
• Exploring the impact of social stress on the adaptive dynamics of COVID-19: Typing the behavior of naïve populations faced with epidemics
• Dealing with complexity of biological systems: from data to models
• Regularization of Mixture Models for Robust Principal Graph Learning
• DeDaL: Cytoscape 3.0 app for producing and morphing data-driven and structure-driven network layouts