Data driven fitting allows you to generate a fit without specifying a parametric equation that describes the relationship between your variables.
Fitit.m is a simple function for data driven curve fitting.
FititDemo.m illustrates how to use fitit to generate a curve fit.
Load_Forecasting.m demonstrates building a short term electricity load (or price) forecasting system with MATLAB. Three non-linear regression models (Boosted Decision Trees, Bagged Decision Trees, and Neural Networks) are calibrated to forecast hourly day-ahead loads given temperature forecasts, holiday information and historical loads. The models are trained on hourly data from the NEPOOL region (courtesy ISO New England) from 2004 to 2007 and tested on out-of-sample data from 2008.
It does not generate partial dependence plots. Is there any in-built matlab tool for that?
Anyone using this "out of the box" should change the line:
cp = cvpartition(100,'k',10);
cp = cvpartition(length(X),'k',10);
how to change the row of data to fit the number of partition in the function cross validation?
It requires the data to be dense and not have large gaps.
The author of this file failed to mention that the number of row of the data to be fit must equals to the number of partition 'N' in the function corss validation. The users have to change it themselves.