search

ealcobaca / pymfe

Python Meta-Feature Extractor package.
https://pymfe.readthedocs.io
MIT License
125 stars 28 forks source link

Support for unsupervised tasks #64

Closed jadermcs closed 4 years ago

jadermcs commented 4 years ago

Is your feature request related to a problem? Please describe. Some metalearning tasks use unsupervised metafeatures (no given y), ex. metastream where in datastreams the target value is not know in some point of time.

Describe the solution you'd like In the MFE().fit(x, y) method, the default value for y could be None so when extracting metafeatures it checks the y to perform a supervised extraction or not. Like scikit-learn implements.

Describe alternatives you've considered The above one is the more concise, an alternative is create another method for extract usupervised features.

Additional context An example in sklearn

FelSiq commented 4 years ago

Hi @jadermcs,

I'm working on your request.

Felipe.

ealcobaca commented 4 years ago

Thank you for suggesting this new feature @jadermcs

lorinczszabolcs commented 4 years ago

Hi @ealcobaca ! Could you give an estimated time when the unsupervised feature will be added? Or is there anything with which we can help?

All the best, Szabi

FelSiq commented 4 years ago

Hi @lorinczszabolcs,

this feature is now on the master branch.

Felipe.

lorinczszabolcs commented 4 years ago

Thank you very much!

Have a nice weekend, Szabi

ealcobaca commented 4 years ago

Thank you @lorinczszabolcs and @FelSiq. You can use this feature by installing our master version. It will be in the PyPI as soon as possible.

Feel free to contribute. We will appreciate new clustering meta-feature. If you wish to contribute, you could check if we miss some clustering meta-feature and provide its implementation to PyMFE.

Best regards :)