Introduces an O(n) evolutionary algorithm claimed to deliver competitive performance for training RBMs with over one million parameters versus CMA-ES and contrastive divergence.
A simple modification in cma-es achieving linear time and space complexity
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NE 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
An Evolutionary Algorithm of Linear complexity: Application to Training of Deep Neural Networks
Introduces an O(n) evolutionary algorithm claimed to deliver competitive performance for training RBMs with over one million parameters versus CMA-ES and contrastive divergence.