Mean field limits for discrete-time dynamical systems via kernel mean embeddings.
Christian FiedlerMichael HertySebastian TrimpePublished in: CoRR (2023)
Keyphrases
- dynamical systems
- linear quadratic
- differential equations
- dynamic systems
- qualitative simulation
- feature space
- kernel methods
- kernel function
- dynamical behavior
- nonlinear dynamical systems
- agent environment
- markov chain
- low dimensional
- euclidean space
- control theory
- markov random field
- linear systems
- markov processes
- phase space
- state space
- bayesian inference
- reproducing kernel hilbert space
- nonlinear dynamics
- predictive state representations
- learning algorithm
- monte carlo
- search algorithm
- dimensionality reduction
- em algorithm