cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar
permonik
Level I

Default setting for Neural Network in JMP14

I would like to know what is the default setting for neural network in JMP14. I am using KFold 10, 3 hidden nodes. I assume that only a single hidden layer and the hyperbolic tangent (TanH) activation function are used. But what about other parameters such as propagation, learning rate, momentum, etc? These are not possible to set for JMP14, but I would like to know what parameters are actually used for those calculation. Thank you.

1 ACCEPTED SOLUTION

Accepted Solutions

Re: Default setting for Neural Network in JMP14

You are correct. You may choose a method of validation and the number of TanH nodes in a single hidden layer. The other parameters are not available. JMP does not use a propogation algorithm to fit the model. The learning rate only applies to boosting.

 

You might find this white paper from JMP helpful to undestand how the Neural platform works.

View solution in original post

3 REPLIES 3
txnelson
Super User

Re: Default setting for Neural Network in JMP14

You may find the information you are looking for in the documentation for Neural Networks
Help0==>Books==>Predictive and Specialized Modeling
Jim
permonik
Level I

Re: Default setting for Neural Network in JMP14

Thank you for your prompt reply, but I checked the manual before contacting JMP community. The information is either not there or I overlooked it. I couldn't find anything about those default settings for non-Pro version of JMP. Just information about a single hidden layer and TanH activation finction.

Re: Default setting for Neural Network in JMP14

You are correct. You may choose a method of validation and the number of TanH nodes in a single hidden layer. The other parameters are not available. JMP does not use a propogation algorithm to fit the model. The learning rate only applies to boosting.

 

You might find this white paper from JMP helpful to undestand how the Neural platform works.