https://communities.sas.com/t5/Statistical-Procedures/Rolling-window-for-forward-scoring-new-dataset-in-rolling-manner/m-p/663797#M31635 <P>Hi,</P><P>&nbsp;</P><P>I am building a decision tree model using proc hpsplit. As I am dealing with time-series data, I want to do a walk-forward validation as suggested instead of 10-fold cross-validation or random sampling as validation set. Validation of the trained decision tree model is done in sliding window:</P><OL><LI>Starting at the beginning of the time series, 10-year window is used to train a model.</LI><LI>The model makes a prediction for the next year (score new data using trained model)</LI><LI>The prediction is stored&nbsp;</LI><LI>Next sliding 10-year window to include the next year and the process is repeated (back to step 1.)</LI></OL><P>How can we do this in SAS?</P><P>Thank you.</P> Sun, 21 Jun 2020 06:31:51 GMT Baolinhdo 2020-06-21T06:31:51Z https://communities.sas.com/t5/Statistical-Procedures/Rolling-window-for-forward-scoring-new-dataset-in-rolling-manner/m-p/663797#M31635 <P>Hi,</P><P>&nbsp;</P><P>I am building a decision tree model using proc hpsplit. As I am dealing with time-series data, I want to do a walk-forward validation as suggested instead of 10-fold cross-validation or random sampling as validation set. Validation of the trained decision tree model is done in sliding window:</P><OL><LI>Starting at the beginning of the time series, 10-year window is used to train a model.</LI><LI>The model makes a prediction for the next year (score new data using trained model)</LI><LI>The prediction is stored&nbsp;</LI><LI>Next sliding 10-year window to include the next year and the process is repeated (back to step 1.)</LI></OL><P>How can we do this in SAS?</P><P>Thank you.</P> Sun, 21 Jun 2020 06:31:51 GMT https://communities.sas.com/t5/Statistical-Procedures/Rolling-window-for-forward-scoring-new-dataset-in-rolling-manner/m-p/663797#M31635 Baolinhdo 2020-06-21T06:31:51Z