No project description provided
- parallel evaluation with joblib (n_jobs)
- implement genetic algorithm to find pipelines
- copy parameters from tpot
- add rules to prevent stupid things (PolynomialFeatures with many columns)
- distribute genetic algorithms with dask
- test joblib distributed backend with dask (nothing to do, just test)
- fine-grained distribution with dask computation graph:
- trivial for prediction
- for fit, each step returns a cross-validated estimate and a fitted model. The fitted model is not used before the final step.
- it is possible to implement cross-validation with a factor 2 improvement when the cross-val and the training folds match
- handle timeouts: https://github.com/dask/distributed/issues/391, https://github.com/dask/dask/issues/1183
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size pamalia-0.0.2-py3-none-any.whl (5.5 kB)||File type Wheel||Python version py3||Upload date||Hashes View|
|Filename, size pamalia-0.0.2.tar.gz (5.2 kB)||File type Source||Python version None||Upload date||Hashes View|