And what do you think about normalization on Neural Network.
I use a package of Mathematica to do some regression with Neural Network algorithm, and I use Feedforward Neural Networks.
In the manual there is written that there are non matter to normalize dataset, but I notice that if I normalize my data between 0 to 1 there is a great improvements of the results.
Normalization of a vector is one of identifying a model by reduction of the number of parameters you need to estimate. An example would be Johansen's normalizatoin of one of the cointegrated vectors in a cointegrated vector autoregression.
Normalization of a vector is one way of identifying a model by reduction of the number of parameters you need to estimate. An example would be Johansen's normalizatoin of one of the cointegrated vectors in a cointegrated vector autoregression.