I want to predict the trend values of a time serie [Y] based on the effect of other 10 input variables that can also have interaction. Since the combination of interaction is unknown, I am applying a regression NN to automatically detect and take that into account. The process I follow is:
Get raw data of [Y] time serie per week
Perfom decomposition and extract the [Y] Trend element per week
Merge observations of the [Y] Trend with the observations of the 10 variables by week into a common table
Normalize (min-max) all variables between 0 and 1
Fit a NN model with Trend as 1 output node and the 10 variables as the input nodes.
Perform k-fold cross validation
Am I following a logic approach or is it anything important that I'm missing?