glemaitre
commented
4 years ago Thanks for the example. However, I think that we have enough synthetic examples. Our priority will be to add more real use-case examples or "anti-pattern" examples. Please refer to the following issue if you want to contribute, we will surely appreciate it:
Describe the issue linked to the documentation
The following notebook is a demo on comparing the performance of three classifiers - SVM, RF and KNN when the input dataset is contaminated with additional noise in higher dimensions of differing variances. This is adapted from the example present in Scikit-learn for classifier comparison - the same input dataset is considered with modifications being the manner in which noise is added to the dataset.
Suggest a potential alternative/fix
The code for the demo for verification and reproduction can be found: https://github.com/sahanasrihari/team-forbidden-forest/blob/master/Sahana/classifier_comparison_noise_dimensions_matplotlib%20(1).ipynb