Bug report : Calibrate (Rescale Confidendes (Logistic)) operator
I wanted to report a bug under certain conditions when AutoModel is executing :
You can reproduce this error by :
- Executing AutoModel with the data in attached file,
- setting the Classification attribute as the target variable.
- setting all the options by default in AutoModel,
After opening the process and investigations :
- The bug is generated by the Calibrate (Rescale Confidences(Logistic)) operator (inside Train Model / Optimize subprocesses) : When this operator is removed (and if also the Split Data operator is removed), the process works fine.
- The bug is linked to the Split Ratio of Train/Test (0.9/0.1). In deed if the ratio is set to 0.8/0.2, the process works fine.
- The bug seems linked to the one-hot-encoded of the Date attributes. In deed if the Extract Date Information is disabled in AutoModel (and thus AutoModel works with the original attributes), the process works fine.
Maybe a possible solution, if the bug is unavoidable under certain conditions, is to use the Calibrate operator with a Handle Exception operator.
Thanks you for your listening,