Universal Time-Uniform Trajectory Approximation for Random Dynamical Systems with Recurrent Neural Networks
This addresses a foundational challenge in machine learning for modeling complex systems, extending beyond previous limitations to compact spaces and finite intervals.
The paper tackles the problem of approximating trajectories of random dynamical systems with recurrent neural networks on non-compact domains over infinite time horizons, achieving a result that allows for any desired accuracy uniformly in time.
The capability of recurrent neural networks to approximate trajectories of a random dynamical system, with random inputs, on non-compact domains, and over an indefinite or infinite time horizon is considered. The main result states that certain random trajectories over an infinite time horizon may be approximated to any desired accuracy, uniformly in time, by a certain class of deep recurrent neural networks, with simple feedback structures. The formulation here contrasts with related literature on this topic, much of which is restricted to compact state spaces and finite time intervals. The model conditions required here are natural, mild, and easy to test, and the proof is very simple.