Skip to content

Latest commit

 

History

History
74 lines (58 loc) · 3.25 KB

plots_nb_YAML_500_(2-250).md

File metadata and controls

74 lines (58 loc) · 3.25 KB

NbClust for YAML-500 (2-250 clusters)

Ball_Hall

  • The mean, through all the clusters, of their mean dispersion.
  • The range is [0,+∞).
  • The difference between two successive slopes to be maximized.

C_Index

  • Shows for a given clustering its fraction of maximal possible increase over minimal distances which within-cluster distances between pairs of points have.
  • The range is [0,1].
  • To be minimized.

Calinski-Harabasz

  • Is proportional to the quotient of the between-group dispersion and pooled within-cluster dispersion.
  • The range is [0,+∞).
  • To be maximized.

McClain_Rao

  • Is the quotient of the mean within-cluster and between-cluster distance.
  • The range is [0,+∞).
  • To be minimized.

Ratkowsky_Lance

  • Is based on the mean of the quotients between between-group dispersion and TSS for each variable of the data.
  • The range is [0,+∞).
  • To be maximized.

Davies-Bouldin

  • Deals with those clusters which are “close” in terms of their barycenters to each other but have very distant points within.
  • The range is [0,+∞).
  • To be minimized.

Silhouette

  • Operates with quantities that only depend on the average distances between a given observation and other observations inside its own and also inside the nearest cluster.
  • The range is [0,1].
  • To be maximized.

Dunn

  • Deals with those clusters which contain the closest points belonging to different clusters and also with clusters that have very distant points within.
  • The range is [0,+∞).
  • To be maximized.