Package pylearn :: Package algorithms :: Module exponential_mean
[hide private]

Module exponential_mean

source code

Modules for maintaining statistics based on exponential decay
Classes [hide private]
ExponentialMean
Maintain an exponentially-decaying estimate of the mean
exp_var
Module with interface similar to ExponentialMean for tracking elementwise variance
DynamicNormalizer
Normalizes input using geometric-decaying estimates of the mean and variance. The output should mean near zero, and variance near 1.
Functions [hide private]
ExponentialMean instance
exp_mean(x, x_shape, max_denom=100)
Return an ExponentialMean to track a Variable x with given shape
source code
ExponentialMean instance
exp_mean_sqr(x, x_shape, max_denom=100)
Return an ExponentialMean to track a Variable x's square with given shape
source code

Imports: copy, numpy, theano


Function Details [hide private]

exp_mean(x, x_shape, max_denom=100)

source code 
Return an ExponentialMean to track a Variable x with given shape
Parameters:
  • x (Variable)
  • max_denom (int)
  • x_shape (tuple)
Returns: ExponentialMean instance

exp_mean_sqr(x, x_shape, max_denom=100)

source code 
Return an ExponentialMean to track a Variable x's square with given shape
Parameters:
  • x (Variable)
  • max_denom (int)
  • x_shape (tuple)
Returns: ExponentialMean instance