Stochastic Gradient Descent (SGD) Algorithm

From GM-RKB
(Redirected from Stochastic Gradient Descent)
Jump to navigation Jump to search

A Stochastic Gradient Descent (SGD) Algorithm is an approximate gradient descent algorithm that is a stochastic optimization algorithm which can be implemented by an SGD System (to solve an SGD task).



References

2018a

2018b

2018c

2015

2014

2012a

2012b

  • http://en.wikipedia.org/wiki/Stochastic_gradient_descent#Background
    • Both statistical estimation and machine learning consider the problem of minimizing an objective function that has the form of a sum: :[math]\displaystyle{ Q(w) = \sum_{i=1}^n Q_i(w), }[/math] where the parameter [math]\displaystyle{ w }[/math] is to be estimated and where typically each summand function [math]\displaystyle{ Q_i( ) }[/math] is associated with the [math]\displaystyle{ i }[/math]-th observation in the data set (used for training).

      In classical statistics, sum-minimization problems arise in least squares and in maximum-likelihood estimation (for independent observations). The general class of estimators that arise as minimizers of sums are called M-estimators. However, in statistics, it has been long recognized that requiring even local minimization is too restrictive for some problems of maximum-likelihood estimation, as shown for example by Thomas Ferguson's example.[1] Therefore, contemporary statistical theorists often consider stationary points of the likelihood function (or zeros of its derivative, the score function, and other estimating equations).

      The sum-minimization problem also arises for empirical risk minimization: In this case, [math]\displaystyle{ Q_i(w) }[/math] is the value of loss function at [math]\displaystyle{ i }[/math]-th example, and [math]\displaystyle{ Q(w) }[/math] is the empirical risk.

      When used to minimize the above function, a standard (or "batch") gradient descent method would perform the following iterations: : [math]\displaystyle{ w := w - \alpha \nabla Q(w) = w - \alpha \sum_{i=1}^n \nabla Q_i(w), }[/math] where [math]\displaystyle{ \alpha }[/math] is a step size (sometimes called the learning rate in machine learning). In many cases, the summand functions have a simple form that enables inexpensive evaluations of the sum-function and the sum gradient. For example, in statistics, one-parameter exponential families allow economical function-evaluations and gradient-evaluations.

      However, in other cases, evaluating the sum-gradient may require expensive evaluations of the gradients from all summand functions. When the training set is enormous and no simple formulas exist, evaluating the sums of gradients becomes very expensive, because evaluating the gradient requires evaluating all the summand functions' gradients. To economize on the computational cost at every iteration, stochastic gradient descent samples a subset of summand functions at every step. This is veryveffective in the case of large-scale machine learning problems.[2]

2010

2007

2006

1992

1952

1951


  1. Ferguson, Thomas S. (1982). "An inconsistent maximum likelihood estimate". J. Am. Stat. Assoc. 77 (380): 831–834. JSTOR 2287314. 
  2. Template:Citation