Login / Signup
Initial state training procedure improves dynamic recurrent networks with time-dependent weights.
Lutz Leistritz
Miroslaw Galicki
Herbert Witte
Eberhard F. Kochs
Published in:
IEEE Trans. Neural Networks (2001)
Keyphrases
</>
initial state
training procedure
recurrent networks
state space
recurrent neural networks
situation calculus
face recognition
linear combination
feed forward
neural network
optimal policy
artificial neural networks
dynamic programming
multi class
principal component analysis
average cost