Max Halford
Max Halford
Hello. I'm opening this issue to make users aware that I've just released a [small package](https://github.com/MaxHalford/pytorch-resample) for sampling from `IterableDataset`s. It's thus complementary to this package which only works with...
I would like to add [Generalized Procrustes Analysis](https://www.wikiwand.com/en/Generalized_Procrustes_analysis) to `prince`. It's quite useful for analyzing free-form surveys. I'll be storing some resources here while I work on it: - [Procrustes...
Hello there! I hope you're doing well. I recently saw this repo pop up and I like it very much. There are some things that we don't yet have in...
- Incremental PCA is already [implemented in scikit-learn](https://scikit-learn.org/stable/auto_examples/decomposition/plot_incremental_pca.html) but seems like it is straightforward to implement. See [this paper](http://cs-www.cs.yale.edu/homes/el327/papers/opca.pdf) too. [This paper](https://arxiv.org/abs/1307.0032) looks quite interesting too. - [Streaming Principal Component...
- It would be called `PerOutputClassifier` - Much like https://scikit-learn.org/stable/modules/generated/sklearn.multioutput.MultiOutputClassifier.html#sklearn.multioutput.MultiOutputClassifier
A nice and simple model selection method from 30 years ago :) https://papers.nips.cc/paper/1993/file/02a32ad2669e6fe298e607fe7cc0e1a0-Paper.pdf
- See [this](https://data-apis.org/dataframe-protocol/latest/index.html) - This way we can make all of our mini-batch code works with all kinds of dataframes
[skrub](https://skrub-data.org/stable/index.html) is a wonderful new project related to scikit-learn. You can see Gaël Varoquaux present it [here](https://www.youtube.com/watch?v=I4Sksd93bZ4). They have a transformer called `GapEncoder`: it's a way to embed fuzzy strings....
See [Wikipedia](https://www.wikiwand.com/en/Savitzky%E2%80%93Golay_filter). The winners of the Kaggle Ashrae competition [used it](https://www.kaggle.com/c/ashrae-energy-prediction/discussion/124709). I checked and it can be computed online. `scipy` [has a function](https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.savgol_coeffs.html) to obtain the convolution coefficients. It seems...