Location: Bunkyo, Japan
Member since: January 2013
Bachelor of Science
Master of Science
The University of Tokyo
Node perturbation learning without noiseless baseline
Enhanced a biological feasibility for node perturbation, a learning rule it might be utilized in song bird.
Optimal node perturbation in linear perceptrons with uncertain eligibility trace
Reduced the residual error in node perturbation learning which might be utilized in song bird brain.