Overview - video

  1. Random project data into lower-dimension space.
  2. Clustering in the lower-dimenstion space with the similarity measurement defined in higher-dimension space.

Similarity Defination