Uploaded February 2021 | Updated September 2026, 10 minutes ago
How to use Kernel-Density Estimation to Group scalars?
We use KDE with a Gaussian kernel, and a log-probability graph to delimit the boundaries of our groups/tiers.
We provide examples and usecases and take a glance at an algorithm proposed to make tierlist of Pokémon players by another vidéaste.
Something that might not be super clear: is that we try to find groups and the red spots at the end represent the limits of the groups. For example, a red dot at point 50, means we have one group below 50 and one group above 50.
Also the audio is a bit bad at the start of the video.
External sources:
youtube.com/watch?v=8yUPhRJtNJM
github.com/jsettlem/elo_world_pokemon_red/blob/master/elo_calculator.py
https://ned.ipac.caltech.edu/level5/March02/Silverman/paper.pdf
scikit-learn.org/stable/modules/density.html#kernel-density
How to use Kernel-Density Estimation to Group scalars?
We use KDE with a Gaussian kernel, and a log-probability graph to delimit the boundaries of our groups/tiers.
We provide examples and usecases and take a glance at an algorithm proposed to make tierlist of Pokémon players by another vidéaste.
Something that might not be super clear: is that we try to find groups and the red spots at the end represent the limits of the groups. For example, a red dot at point 50, means we have one group below 50 and one group above 50.
Also the audio is a bit bad at the start of the video.
External sources:
youtube.com/watch?v=8yUPhRJtNJM
github.com/jsettlem/elo_world_pokemon_red/blob/master/elo_calculator.py
https://ned.ipac.caltech.edu/level5/March02/Silverman/paper.pdf
scikit-learn.org/stable/modules/density.html#kernel-density
