Neural networks don't have tidy formulas like regression. If you really want to see the nn model, you would have to look at the xml file produced when you export the model, but it wouildn't be easy to convert that into an Excel calculator.
--
Original Message:
Sent: 9/11/2023 7:38:00 AM
From: Mehmet Muzaffer
Subject: RE: SPSS 26: What is the difference between the regression and neural method regarding their usage for prediction?
I have the same question. I would like to obtain the MLP equation for my study, because I would like to provide a calculator (an excel calculator for example). There has to be some kind of equation since I can apply this xml file to another dataset. But I don't know how to get it.
------------------------------
Mehmet Muzaffer
------------------------------
Original Message:
Sent: Mon December 26, 2022 01:55 AM
From: jamal numan
Subject: SPSS 26: What is the difference between the regression and neural method regarding their usage for prediction?
Not sure regarding the exact difference between the regression and neural network. Do you mean that there is no statistical (mathematical) equation that relates the variables in order to produce the predicted value?
In this sense, what could be the difference between the mathematical and statistical equations?
------------------------------
jamal numan
Original Message:
Sent: Thu December 22, 2022 10:09 AM
From: Jon Peck
Subject: SPSS 26: What is the difference between the regression and neural method regarding their usage for prediction?
NN models do not provide statistical results such as significance levels and, in fact, there isn't a model there in the same sense as there is in linear (or nonlinear) regression.
--
Original Message:
Sent: 12/22/2022 8:33:00 AM
From: jamal numan
Subject: SPSS 26: What is the difference between the regression and neural method regarding their usage for prediction?
SPSS 26: What is the difference between the regression and neural method regarding their usage for prediction?
I'm not sure why the regression method is considered to be statistical while the neural is deemed as machine learning!
- Both are constructed based on a number of given observations
- Both are optimizing the error (residual) between the observed value (actual) and the predicted value
- Both are trying to find the weight of each variable that achieves the minimal residual
Then how to make a distinction between the statistical, mathematical, and machine learning methods?
------------------------------
jamal numan
------------------------------
#SPSSStatistics