Framing regression as an actor-critic reinforcement learning problem with a Gaussian reward can fit a noisy sine wave, but the demonstration reduces to squared-error minimization and adds no new method.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RL as Regressor: A Reinforcement Learning Approach for Function Approximation
Framing regression as an actor-critic reinforcement learning problem with a Gaussian reward can fit a noisy sine wave, but the demonstration reduces to squared-error minimization and adds no new method.