Using Forecast Data in Optimised Classification Models to make Predictions, and also determine accuracy or probability
1 view (last 30 days)
Hi I am working on a project that will use some historic data, 1 year approximately, to determine some trading strategies based on certain variables. I have currently created code that makes classification models based on this data, and I would like to use these ready-made models with some forecast data to make the same predictions to determine Trading Strategies.
I am currently using fitcsvm, fitctree, fitcensemble and fitcnb to make the models.
The File attached WindQuestion2 is the file in question, with connection strings and filenames taken out, so don't worry about warning/error lines around fopen or whatever. I also get the data through Analysis Services, but that shouldn't make much of a difference.
If it is possible, I want to use the data in FCStruct (approx line 220), in the models that are created in lines 82-98. So that I can make predictions on what strategy should be better, also is it possible to have some probability attached?
I have tried to make a confusion matrix in the same vein, but I haven't been able due to being on 2017a.