Tpot: Custom Validation Set for tpot.fit

[skrish13]

How do i give a seperate validation set so that tpot evaluates pipelines using that instead of random split of trainset itself?

[weixuanfu]

Please check the cv parameter in TPOT API. You can merge the validation set with the trainset for fitting in tpot_obj.fit(X, y) and then specify train/test splits via an iterable (see the example for GridSearchCV)

[sahil2897]

Can you please explain a little more how to custom specify training and validation data sets. I don't want to use k-fold cv?

[weixuanfu]

You could try other split models like cv=TimeSeriesSplit(n_splits=5) or use indexes of samples based on the example for GrifSearchCV above.