|
Activation Function Scale dialog box |
|
Use this dialog box to scale the activation function of the output layer in a neural network. Typically, one would do this to match the extent of the output of a neural network to the extent of target values.
Natural Function Range
Displays the minimum value of the currently set activation function as a reference.
Displays the maximum value of the currently set activation function as a reference.
 | The name of the currently set activation function is displayed on the title bar. |
Desired Output Range
Indicates the minimum value that the currently set activation function will evaluate to under the current scaling. Type in this box the minimum value you want to allow this function to reach.
Indicates the maximum value that the currently set activation function will evaluate to under the current scaling. Type in this box the maximum value you want to allow this function to reach.
| It is a good practice to specify a certain amount of leeway between the extent of the activation function and the target values as to allow neural network outputs to reach unrealized values outside the range of the sample at hand. |
Click this button to set the Min. Value and the Max. Value boxes to the extent of the target values. Predictive Systems Lab will carry out a scan of the target values to determine their range.
Click this button to restore the Min. Value and the Max. Value boxes to their previously set values.
Click this button to restore the Min. Value and the Max. Value boxes to the natural activation function range.
Click this button to close the dialog box and save any changes you have made.
Click this button to close the dialog box without saving any changes you have made.
Sponsored
Try Predictive Systems Lab
Need Help?
Contact support