Unsupervised learning portion of the project

[SBuxton-IDM]

This issue is meant to organize several different issues relating to the unsupervised learning portion of the project. As the project progresses I may add more issues/checkboxes

• Conduct PCA on existing dataset
• Compare to what we think are the 5 key features (Download mbps, latency, jitter, upload etc.)
• Create a visualization (like heat map or biplot) for the results
• Define a variable in the notebook that contains only the key features (which we will use to train our model)

[SBuxton-IDM]

Blocked by: #5

[yolanda707]

PCA maybe not enough, I am not sure the relationship of data is linear or non-linear, I plan to use UMAP, PCA and t-SNE to check at the same time. But still waiting for the one complete dataset to work on, now we just have data in 4 different folders which is not working for the project.

[yolanda707]

Have uploaded the UL procession file for checking the steps of unsupervised learning deduction dimension and k-means to get the clustering. Hope to have your review comments here.

[SBuxton-IDM]

Have uploaded the UL procession file for checking the steps of unsupervised learning deduction dimension and k-means to get the clustering. Hope to have your review comments here.

Looks great! I'm assuming the stuff after a certain line (mentioning food) is for another project, but this fulfills all the requirements detailed in the issue, I especially like the added 3d point plot. I will make a PR on Saturday to make this presentation-ready and add you as a reviewer (just getting rid of the extra cells and adding some markdown cells for legibility).