RBF NN using newrb for prediction one target "output vector" from many inputs factors " 15 vectors"

 I am working in a research for forecasting electricity price. I am using a RBF network "newrb" for predicting electricity price . the network has 15 inputs (different factors) and one output(electricity price). i am asking how to do that using newrb for different Input & Output matrices ( 15 input predictors for one Output).

net = newrb(P,T,goal,spread,MN,DF).

P: R-by-Q matrix of Q input vectors.

T: S-by-Q matrix of Q target class vectors.

in my case P matrix ≠ T matrix

[srikanth allamsetty]

Hi, I am also working on a similar kind of problem but in a different area, with 5 input parameters and one output, which need to be predicted. Did you find any way to solve your problem using RBF? if so please provide me some information. Thank you in advance.

Answer 1

[Greg Heath]

Search the NEWSGROUP and ANSWERS using

 greg newrb

Concentrate on the ones used for classification and pattern-recognition.

Hope this helps.

Thank you for formally accepting my answer

Greg

[Abdussalam]

Thank you Greg, I have seen your contribution in some questions before i ask mine. I searched too as you recommend in your answer, i found them not related to my case. I am not working on classification as many questions are, i am working on prediction one output from 15 input predictors. I used NN fitting tool and it works fine. but my question is can that be done with RBF "newrb" or not?

[Greg Heath]

newrb works just as well for regression.

The biggest problems with newrb are

 1.  You cannot assign an initial configuration
 2.  It does not have the flexibility of newfit or patternnet.