TABLE 2.

Learning simulation parameters and results. True and learned double-gyre model parameters over 10 learning episodes.

Learning simulation parameters
Number of episodes	10 with 10 steps each
Tile coding	8 tilings with 8 × 8 tiles each
Step size α	0.9
Trace decay rate λ	0.9
ɛ-greedy parameter	0.15
Double-Gyre Model Parameters and Results
True μ	0.25
Learned μ	0.2481
True ω dg	π/5 ≈ 0.6283
Learned ω dg	0.6344