110,000 UMAP

• Version of Opinion cluster analysis of 110,000 people using UMAP instead of PCA

Impression "Like a stingray"

• I'm curious if the tail part is a political feature or a systemic one (e.g., people choosing "I don't know" or "neutral" on everything).

  • If you look at the data, there's a "don't know" for each question, and I'm including that one-hot expression.

• If even UMAP can't separate them, there's no way PCA can.

  • There was some kind of cluster sticking out of the PCA, so maybe that's the stingray's tail.

• The opinion vectors of anonymous citizens are not so clearly separated.

This page is auto-translated from /nishio/11万人のUMAP using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.