3 problems

0 votes0 replies1 view
Boundedness of the proxy approximation between TD error and value gradients
Boundedness of the proxy approximation. There exists a constant $C>0$ such that, for a sufficiently fine state-space discretization,
Conjectureopenreinforcement-learning stochastic-control temporal-difference-learningarXiv posted 2 days ago
0 votes0 replies0 views
Extension of the asymptotic bias formula to general eligibility traces
The TD learning recursion is governed by a Markov process, with eligibility-trace parameter $\lambda$ ; for $\lambda=0$ , the asymptotic bias limit exists and is given by the bias fo…
Conjectureopenreinforcement-learning stochastic-approximation temporal-difference-learningarXiv posted 2 days ago
0 votes0 replies0 views
Extension of federated TD analysis to error-feedback encoding and realistic channels
The Quantized Federated TD learning algorithm, or QFedTD, updates the server parameter by … where $\alpha$ is a constant step size, $b_{i,k}$ are independent Bernoulli packet-deliv…
Conjectureopencommunication-channels federated-learning temporal-difference-learningarXiv posted 2 days ago

Boundedness of the proxy approximation between TD error and value gradients