I have had zero luck with the TImeSeries modeling tools so far. I know that docs discusses an issue with the optima maxima (sp??) - maybe that's my issue?? I didn't quite understand what the docs were telling my regarding that but it sounds like something I needed to avoid in my modeling.
So instead I've started experimenting with Logistic Regression using 1's and 0's for BUY AND SELL signals for nextdayclose_percent change. I like this better because I can include my own indicators that I know have strong correlations to the data - something I can't do with the Time Series model which only allows a single column of data as inputs....
These forums are a nice addition for us users -. I'm finding reading other people most to be a good source of insight... thank you for adding them Gepsoft guys (and girls :>)
It looks like you're new here. If you want to get involved, click one of these buttons!