*'a generic approach to minimizing R(θ) (the cost function) by gradient descent.'*

Given a neural network with inputs X with hidden layers comprised of hidden units Z used to predict some target T, we can represent a neural network schematically (simplifying the notation in Hastie by omitting key subscripts and summations)

X -> Z -> T

Z = σ( α

_{0}+ α

^{T}x)

T = β

_{0}+ βZ

f(X) = g(T) [1]

where σ = the activation function

Given weights {α

_{0},α

_{0}, β

_{0}, β} find the values that minimize the specified error function:

R(

*θ) =*∑∑ ( y-f(x)

^{2}

_{ }) [2] (note a number of possible error functions may be used)

Backpropogation equations:

s = σ'( α

Gradient Descent Update:

^{T}x )βδ [3]
Errors can be re-specified as:

∂R/ ∂β = δZ [4]

∂R/ ∂α = sx [5]

Gradient Descent Update:

β

^{r+1}= β^{r}- γ ∂R/ ∂β [6]
α

^{r+1}= α^{r}- γ ∂R/ ∂α [7]**Algorithm:**

*Forward Pass:*use initial or current weights (guesses) and calculate f(X), and errors δ from the output layer [2]

*Backward Pass:*'back propagate' via back propagation equation [3] to obtain s. Both sets of errors (δ) and (s) are used to derive the derivative terms in [4] and [5] which are then used in the gradient descent update weight estimates via equations [6]& [7].

In Predictive modeling with SAS Enterprise Miner by Sarma, the following basic description of back propagation is given:

Specify an error function E.

1) 1st iteration- set initial weights, use to evaluate E

2) 2nd iteration- weights are changed by a small amount such that the error is redced

-repeat until convergence

As Sarma explains, with each iteration a number of weights are produced, so if it takes 100 iterations to converge, 100 possible models are specified, giving 100 sets of weights. Using validation data, the best iteration can be chosen calculating E via the validation data.

## No comments:

## Post a Comment

Note: Only a member of this blog may post a comment.