auto-sklearn
auto-sklearn copied to clipboard
automl
Automated Machine Learning with scikit-learn
Any plans on adding an AutoSklearn2Regressor based on the latest paper?
I'm trying to extend autosklearn estimators list with XGBoostClassifier. When XGBoostClassifier runs alone (as an estimator), it uses roughly half of the available cores (and appears to be working just...
I would like to extend auto-sklearn to handle datasets with both numerical and textual features. In particular, I want to implement a custom preprocessor that can take a textual feature...
Some ideas on how to improve the readme and docs: * [Readme] Add pipeline image * ~~[Readme] Make publications bibtext entries that can be easily copied~~ * [Readme] Add further...
## Describe the bug ## If all evaluations crash, auto-sklearn does not even show a warning ## To Reproduce ## For example add a new model class which always crashes...
Previously, we used to delete the backend with `__del__` in `AutoML` but this was changed with PR #1155 due to issues of modules being unloaded before `__del__` had finished. The...
From issue #1302, it appears autosklearn is a bit unstable when run many times in the same script, i.e. in a for loop. ```python for i in range(400): automodel =...
`PIPELINE_DATA_DTYPE` is currently outdated, we now accept pandas dataframes, need to update pipeline components to ensure this is supported throughout and update the type to reflect this. This was indentified...
Currently the test code has a lot of creating objects just to get the configuration space and then setting hyperparameters. ```python pipeline = SimpleRegressionPipeline(dataset_properties=dataset_properties) cs = pipeline.get_hyperparameter_search_space() # ...later config...
It appears the the code still accepts `data_preprocessor` as a valid entry to `include: Dict[str, Any]` of the estimators. It states the only valid entry is `'feature_type'` if passing `data_preprocessor:...