Forecasting Chaotic Systems with Very Low Connectivity Reservoir Computers
| Autor: | Daniel J. Gauthier, Aaron Griffith, Andrew Pomerance |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Mathematical optimization Computer Science - Machine Learning Computer science Chaotic MathematicsofComputing_NUMERICALANALYSIS General Physics and Astronomy FOS: Physical sciences Machine Learning (stat.ML) 01 natural sciences Measure (mathematics) Field (computer science) 010305 fluids & plasmas Machine Learning (cs.LG) Statistics - Machine Learning 0103 physical sciences Attractor 010306 general physics Mathematical Physics Hyperparameter Structure (mathematical logic) Applied Mathematics Bayesian optimization Statistical and Nonlinear Physics Nonlinear Sciences - Chaotic Dynamics Chaotic Dynamics (nlin.CD) Heuristics |
| DOI: | 10.48550/arxiv.1910.00659 |
| Popis: | We explore the hyperparameter space of reservoir computers used for forecasting of the chaotic Lorenz '63 attractor with Bayesian optimization. We use a new measure of reservoir performance, designed to emphasize learning the global climate of the forecasted system rather than short-term prediction. We find that optimizing over this measure more quickly excludes reservoirs that fail to reproduce the climate. The results of optimization are surprising: the optimized parameters often specify a reservoir network with very low connectivity. Inspired by this observation, we explore reservoir designs with even simpler structure, and find well-performing reservoirs that have zero spectral radius and no recurrence. These simple reservoirs provide counterexamples to widely used heuristics in the field, and may be useful for hardware implementations of reservoir computers. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|