Remote Sensing Landslide Recognition Based on Convolutional Neural Network
Algorithm 1
Adam pseudocode. Adam is our proposed algorithm for stochastic optimization (see Section 2 for details and for a slightly more efficient (but less clear) order of computation). indicates the elementwise square . Good default settings for the tested machine learning problems are ,,, and . All vector operations are elementwise. With and , we denote and to the power .
Require:: Step size
Require:: Exponential decay rates for moment estimates
Require:: Stochastic objective function with parameters
Require:: Initial parameter vector
(Initialize moment vector)
(Initialize moment vector)
(Initialize timestep)
while not converged, do
(Get gradients w.r.t. stochastic objective at timestep )
(Update biased first moment estimate)
(Update biased second raw moment estimate)
(Compute bias-corrected first moment estimate)
(Compute bias-corrected second raw moment estimate)