Minimizing Power for Neural Network Training with Logarithm-Approximate Floating-Point Multiplier.

TaiYu Cheng Jaehoon Yu Masanori Hashimoto

Published in: PATMOS (2019)

Keyphrases

floating point
neural network training
neural network
training algorithm
fixed point
floating point unit
sparse matrices
instruction set
floating point arithmetic
optimization method
particle swarm optimisation
decision trees
genetic programming
control system
search space
similarity measure
learning algorithm