Learning Interaction Kernels for Agent Systems on Riemannian Manifolds

Interacting agent and particle systems are extensively used to model complex phenomena in science and engineering. We consider the problem of learning interaction kernels in these dynamical systems constrained to evolve on Riemannian manifolds from given trajectory data. The models we consider are based on interaction kernels depending on pairwise Riemannian distances between agents, with agents interacting locally along the direction of the shortest geodesic connecting them. We show that our estimators converge at a rate that is independent of the dimension of the state space, and derive bounds on the trajectory estimation error, on the manifold, between the observed and estimated dynamics. We demonstrate the performance of our estimator on two classical first order interacting systems: Opinion Dynamics and a Predator-Swarm system, with each system constrained on two prototypical manifolds, the 2-dimensional sphere and the Poincaré disk model of hyperbolic space.

[1]  Shihao Yang,et al.  Inference of dynamic systems from noisy and sparse data via manifold-constrained Gaussian processes , 2021, Proceedings of the National Academy of Sciences.

[2]  Fei Lu,et al.  Learning interaction kernels in heterogeneous systems of agents from multiple trajectories , 2019, J. Mach. Learn. Res..

[3]  Hao Liu,et al.  Doubly Robust Off-Policy Learning on Low-Dimensional Manifolds by Deep Neural Networks , 2020, ArXiv.

[4]  S. Brunton,et al.  Discovering governing equations from data by sparse identification of nonlinear dynamical systems , 2015, Proceedings of the National Academy of Sciences.

[5]  B. Piccoli,et al.  A nonlinear model of opinion formation on the sphere , 2015 .

[6]  John K. Tomfohr,et al.  Lecture Notes on Physics , 1879, Nature.

[7]  T. Kolokolnikov,et al.  A minimal model of predator–swarm interactions , 2014, Journal of The Royal Society Interface.

[8]  S. Strogatz.  From Kuramoto to Crawford: exploring the onset of synchronization in populations of coupled oscillators , 2000 .

[9]  W. Ashby,et al.  Principles of the self-organizing dynamic system. , 1947, The Journal of general psychology.

[10]  Steven L. Brunton,et al.  Data-driven discovery of partial differential equations , 2016, Science Advances.

[11]  R. Stephenson.  A and V , 1962, The British journal of ophthalmology.

[12]  Adam Krzyzak,et al.  A Distribution-Free Theory of Nonparametric Regression , 2002, Springer series in statistics.

[13]  Nastassia Pouradier Duteil,et al.  Opinion dynamics on a general compact Riemannian manifold , 2017, Networks Heterog. Media.

[14]  Christian Soize,et al.  Probabilistic learning on manifolds constrained by nonlinear partial differential equations for small datasets , 2020, Computer Methods in Applied Mechanics and Engineering.

[15]  Felipe Cucker,et al.  On the mathematical foundations of learning , 2001 .

[16]  R. Fetecau,et al.  Self-organization on Riemannian manifolds , 2018, Journal of Geometric Mechanics.

[17]  Mariela Morveli Espinoza,et al.  Multi-Agent Systems based on Contextual Defeasible Logic considering Focus , 2020, ArXiv.

[18]  E. Hairer,et al.  Geometric Numerical Integration: Structure Preserving Algorithms for Ordinary Differential Equations , 2004 .

[19]  M. Maggioni,et al.  On the identifiability of interaction functions in systems of interacting particles , 2019, Stochastic Processes and their Applications.

[20]  G. Karniadakis,et al.  Multistep Neural Networks for Data-driven Discovery of Nonlinear Dynamical Systems , 2018, 1801.01236.

[21]  Mauro Maggioni,et al.  Learning Interaction Kernels in Stochastic Systems of Interacting Particles from Multiple Trajectories , 2020, Foundations of Computational Mathematics.

[22]  Seung‐Yeal Ha,et al.  Emergent behaviors of Cucker–Smale flocks on the hyperboloid , 2020, 2007.02556.

[23]  E. Tadmor,et al.  From particle to kinetic and hydrodynamic descriptions of flocking , 2008, 0806.2182.

[24]  Ernst Hairer,et al.  Geometric Integration of Ordinary Differential Equations on Manifolds , 2001 .

[25]  John M. Lee.  Introduction to Smooth Manifolds , 2002 .

[26]  Craig W. Reynolds.  Flocks, herds, and schools: a distributed behavioral model , 1987, SIGGRAPH.

[27]  Mark Braverman,et al.  On the convergence of the Hegselmann-Krause system , 2012, ITCS '13.

[28]  Giang Tran,et al.  Exact Recovery of Chaotic Systems from Highly Corrupted Data , 2016, Multiscale Model. Simul..

[29]  Aleksander Byrski,et al.  Evolutionary Multi-Agent Computing in Inverse Problems , 2013, Comput. Sci..

[30]  Arend Hintze,et al.  Exploring the coevolution of predator and prey morphology and behavior , 2016, ALIFE.

[31]  Mauro Maggioni,et al.  Learning theory for inferring interaction kernels in second-order interacting agent systems , 2020, Sampling Theory, Signal Processing, and Data Analysis.

[32]  Ming Zhong,et al.  Nonparametric inference of interaction laws in systems of agents from trajectory data , 2018, Proceedings of the National Academy of Sciences.

[33]  Tiangang Cui,et al.  Data‐driven model reduction for the Bayesian solution of inverse problems , 2014, 1403.4290.

[34]  U. Krause.  A DISCRETE NONLINEAR AND NON–AUTONOMOUS MODEL OF CONSENSUS FORMATION , 2007 .

[35]  M. Fornasier,et al.  Inferring Interaction Rules From Observations of Evolutive Systems I: The Variational Approach , 2016, 1602.00342.

[36]  Qiang Du,et al.  Discovery of Dynamics Using Linear Multistep Methods , 2019, SIAM J. Numer. Anal..

[37]  Mauro Maggioni,et al.  Data-driven Discovery of Emergent Behaviors in Collective Dynamics , 2020, Physica D. Nonlinear phenomena.

[38]  Guang Lin,et al.  Robust data-driven discovery of governing physical laws with error bars , 2018, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[39]  Guillaume Deffuant,et al.  Interacting Agents and Continuous Opinions Dynamics , 2001, cond-mat/0111494.

[40]  I. Couzin,et al.  Inferring the structure and dynamics of interactions in schooling fish , 2011, Proceedings of the National Academy of Sciences.

[41]  Jadran Vrabec,et al.  Lennard-Jones force field parameters for cyclic alkanes from cyclopropane to cyclohexane , 2015 .

[42]  Boundary-interior principle for microbial navigation in geometric confinement , 2020, 2011.02811.

[43]  Daniele Nardi,et al.  DOP: Deep Optimistic Planning with Approximate Value Function Evaluation , 2018, AAMAS.

[44]  Leah Edelstein-Keshet,et al.  Inferring individual rules from collective behavior , 2010, Proceedings of the National Academy of Sciences.

[45]  S. Osher,et al.  Sparse dynamics for partial differential equations , 2012, Proceedings of the National Academy of Sciences.

[46]  Alain Sarlette,et al.  Consensus Optimization on Manifolds , 2008, SIAM J. Control. Optim..

[47]  Steven L. Brunton,et al.  Deep learning of dynamics and signal-noise decomposition with time-stepping constraints , 2018, J. Comput. Phys..