Login / Signup
Shallow Network Training With Dynamic Sample Weights Decay - a Potential Function Approximator for Reinforcement Learning.
Leo Ghignone
Michael Barlow
Published in:
SSCI (2019)
Keyphrases
</>
potential functions
reinforcement learning
weighted sums
markov random field
network structure
maximum entropy
pairwise
power law
reinforcement learning algorithms
hidden nodes
information retrieval
learning algorithm
training samples
function approximation
function approximators
iterative conditional