Comparing the performance of different clustering algorithms on toy datasets on adding high dimensional gaussian noise

[sree0917]

Aim: Analyzing how the performance of different clustering algorithms for different datasets change on adding noise with different dimensions:

This demo is a Jupyter Notebook documentation describing the effect of the addition of different dimensions of noise on a dataset. Here different types of synthetic datasets are generated on which the experiment is performed. To these datasets Gaussian noise of different dimensions are added, and the performance of each clustering algorithm is measured after noise addition. This is repeated for noise with different variances.

Output: The plots that compare the effect of varying noise dimensions on different clustering algorithms for each of the datasets. In this set of subplots, the variance of the added noise changes along the column and the dataset changes along the row.

Link to the demo: https://nbviewer.jupyter.org/github/sree0917/scikit-learn/blob/master/clustering_comparison_pr.ipynb

[bdpedigo]

• when you plot the data, remove ticks, tick labels, and have only one legend to the right of the entire figure (rather than 3). This will give you a lot more space. And can then make the titles bigger.
• python should use snake_case (can look this up) as a convention for naming variables. For example addNoise should be add_noise. Many other variables like this should be changed
• I think the last plot is the only ARI one that we need (so that's all I'm going to give feedback on)
• can you add the datasets plot on top of the last ARI plot, and then remove the datasets alone plot?
• is the variance of DBSCAN always 0? That seems weird to me, but I also don't know much about DBSCAN
• not sure I agree with your claim about spectral clustering winning

Overall, nice plots and nice work! These are fairly minor comments.

[sree0917]

Thanks for the feedback. I have made necessary changes.

[bdpedigo]

Unclear what file is the right one now? The notebook in the link at the top looks the same. Are you trying to merge two files?

[bdpedigo]

also this PR should be into the neurodatadesign fork

[bdpedigo]

seems like there are still plotting things that you didn't change, if I am looking at the right notebook

[sree0917]

Hi! I have made the necessary changes and made a new PR into the NeuroDataDesign fork.
Link to the PR - NeuroDataDesign@0a71e4d

[sree0917]

Hi! I have made PR the right way now.

Link - NeuroDataDesign#25