Quote:
Originally Posted by MJC922
I'm not familiar with what you're using for files or a database. If you can post a csv file I might be able to look at it and then post some basic example. I can't post my own stuff because there's too much in it that I'd rather not bring visibility to. There are several ways of using solver and I highly doubt that I use it as elegantly as some do but might be able to get you thinking about various options. I find there's actually a lot data preparation that I need to make before I can even think about using solver on the data. Some of that can be done in excel, some in access. I don't know if you're familiar with the term discretization? Anything like rankings within the race itself or binary characteristics etc etc, there's truly a myriad number of ways to prepare the data that you have. Solver is all about optimization, not really regression, you're optimizing on win rate, or optimize on ROI, whatever your goal is for what you're looking at. The other thing you need is a big sample, split it, solve on the first sample and when it crushes the sample with a +50% ROI you can calm yourself down real quick when it loses more than takeout on the other sample. **Edit I'm not sure what type of file would be acceptable to post out here in the first place so disregard my request to post any type of CSV file if it violates the rules of the board.
|
thank you but what exactly you need....
say i have factor values for fr1 fr2 fr3 and finish position for 1000 horses
is that enough?
