Skip to main content
Bumped by Community user
Bumped by Community user
Bumped by Community user
Post Reopened by Sycorax neural-networks
By request, clarified question. Previously it sounded like a duplicate.
Source Link
RMurphy
  • 876
  • 7
  • 18

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in aIt is already common to do something "like"**(see asterisks below) z-standardization of the outputs of one neural network. However, web searching this leads layer before passing it to the next. z-standardization would transform the columns of $H_{\ell}W_{\ell} + \beta_{\ell}$ (where $\ell$ denotes a delugelayer and $H$ denotes a "hidden" matrix containing the values of whether batch norm should be beforehidden neurons, or after an activation which is not my concernthe input data) to have 0 mean and unit standard deviation.

Despite being marked as a duplicate**In reality, thisbatchnorm is not a duplicate, please see my comment below.

Moreoverused, we can simplifywhich incorporates learnable weights to the discussionstandardization function to justhelp the model "undo" or "modify" the deterministic nature of z-standardization; that is, centering to mean 0 and scaling to unit standard deviationscaling.

(1) I observe that z-scaling should bescoring is a nonlinear function in $W_{\ell}$, because we must compute the weightssample standard deviation of the layers that come before it$H_{\ell}W_{\ell} + \beta_{\ell}$, due to square andwhich involves a square root operations. It follows that batch-norm will be nonlinear in the previous layer's weights as well.

A motivation for using nonlinearities is that they(2) Thus, if we do batch norm, we do not need to use a common activation function such as ReLU or tanh to prevent the whole composition"stack" (composition) of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this Meanwhile, the community generally believes batch norm is "good". Thus, why not just use batch norm between layers and free ourselves from choosing between ReLU, ELU, etc. etc.?

The discussion generalizes to batch-norm^^ that is the purpose of my question

what follows is an observation of why my question might also be useful:

(3) Then, where learnable scalings and centerings are employedwe may observe all the questions online about whether batch norm should be used before or after the activation function. But couldn't we just use batch norm and avoid this question?

Insights and/or references on this specific topic would be much appreciated!

Thanks

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in a neural network. However, web searching this leads to a deluge of whether batch norm should be before or after an activation which is not my concern.

Despite being marked as a duplicate, this is not a duplicate, please see my comment below.

Moreover, we can simplify the discussion to just z-standardization; that is, centering to mean 0 and scaling to unit standard deviation.

z-scaling should be nonlinear in the weights of the layers that come before it, due to square and square root operations.

A motivation for using nonlinearities is that they prevent the whole composition of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this?

The discussion generalizes to batch-norm, where learnable scalings and centerings are employed.

Insights and/or references on this specific topic would be much appreciated!

It is already common to do something "like"**(see asterisks below) z-standardization of the outputs of one neural network layer before passing it to the next. z-standardization would transform the columns of $H_{\ell}W_{\ell} + \beta_{\ell}$ (where $\ell$ denotes a layer and $H$ denotes a "hidden" matrix containing the values of hidden neurons, or the input data) to have 0 mean and unit standard deviation.

**In reality, batchnorm is used, which incorporates learnable weights to the standardization function to help the model "undo" or "modify" the deterministic nature of z-scaling.

(1) I observe that z-scoring is a nonlinear function in $W_{\ell}$, because we must compute the sample standard deviation of $H_{\ell}W_{\ell} + \beta_{\ell}$, which involves a square root. It follows that batch-norm will be nonlinear in the previous layer's weights as well.

(2) Thus, if we do batch norm, we do not need to use a common activation function such as ReLU or tanh to prevent the whole "stack" (composition) of affine layers from collapsing into one affine layer. Meanwhile, the community generally believes batch norm is "good". Thus, why not just use batch norm between layers and free ourselves from choosing between ReLU, ELU, etc. etc.?

^^ that is the purpose of my question

what follows is an observation of why my question might also be useful:

(3) Then, we may observe all the questions online about whether batch norm should be used before or after the activation function. But couldn't we just use batch norm and avoid this question?

Insights and/or references on this specific topic would be much appreciated!

Thanks

added 94 characters in body
Source Link
RMurphy
  • 876
  • 7
  • 18

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in a neural network. However, web searching this leads to a deluge of whether batch norm should be before or after an activation which is not my concern.

Despite being marked as a duplicate, this is not a duplicate, please see my comment below.

Moreover, we can simplify the discussion to just z-standardization; that is, centering to mean 0 and scaling to unit standard deviation.

z-scaling should be nonlinear in the weights of the layers that come before it, due to square and square root operations.

A motivation for using nonlinearities is that they prevent the whole composition of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this?

The discussion generalizes to batch-norm, where learnable scalings and centerings are employed.

Insights and/or references on this specific topic would be much appreciated!

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in a neural network. However, web searching this leads to a deluge of whether batch norm should be before or after an activation which is not my concern.

Moreover, we can simplify the discussion to just z-standardization; that is, centering to mean 0 and scaling to unit standard deviation.

z-scaling should be nonlinear in the weights of the layers that come before it, due to square and square root operations.

A motivation for using nonlinearities is that they prevent the whole composition of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this?

The discussion generalizes to batch-norm, where learnable scalings and centerings are employed.

Insights and/or references on this specific topic would be much appreciated!

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in a neural network. However, web searching this leads to a deluge of whether batch norm should be before or after an activation which is not my concern.

Despite being marked as a duplicate, this is not a duplicate, please see my comment below.

Moreover, we can simplify the discussion to just z-standardization; that is, centering to mean 0 and scaling to unit standard deviation.

z-scaling should be nonlinear in the weights of the layers that come before it, due to square and square root operations.

A motivation for using nonlinearities is that they prevent the whole composition of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this?

The discussion generalizes to batch-norm, where learnable scalings and centerings are employed.

Insights and/or references on this specific topic would be much appreciated!

Post Closed as "Duplicate" by Sycorax neural-networks
Source Link
RMurphy
  • 876
  • 7
  • 18

Batch Normalization or just z-normalization as a Nonlinearity

Reason for new question: I have a question about batch normalization as well as nonlinear "activation functions" in a neural network. However, web searching this leads to a deluge of whether batch norm should be before or after an activation which is not my concern.

Moreover, we can simplify the discussion to just z-standardization; that is, centering to mean 0 and scaling to unit standard deviation.

z-scaling should be nonlinear in the weights of the layers that come before it, due to square and square root operations.

A motivation for using nonlinearities is that they prevent the whole composition of affine layers from collapsing into one affine layer. Isn't standardization enough to avoid this?

The discussion generalizes to batch-norm, where learnable scalings and centerings are employed.

Insights and/or references on this specific topic would be much appreciated!