andremarco/kde-and-clustering-on-running-data

KDE-and-clustering-on-running-data

To properly visualize the code, choice the raw data visualization of the .Rmd file

Part 1

Estimate and visualize properly and meaningfully the density of latitude and longitude data (Kernel Density Estimation) Figure out a way to single out the places where the runner run the most.

Part 2

Functional data analysis Pick a possibly meaningful epsilon (and/or play around with different epsilon ’s). With this choice:

• find the top-5 paths with the highest local density,
• find the low-5 paths with the lowest local density,
• now that you have a kernel estimator, you can run mean-shift and cluster the tracks