Class PROCLUS

constructor

• new PROCLUS(
      k: number,
      a: number,
      b: number,
      l: number,
      minDeviation?: number,
  ): PROCLUS

  Parameters

  • k: number

    Number of clusters
  • a: number

    Number to multiply the number of clusters for sample size
  • b: number

    Number to multiply the number of clusters for final set size
  • l: number

    Average dimensions
  • OptionalminDeviation: number

    Minimum deviation to check the medoid is bad

  Returns PROCLUS

_a

_b

_bestObjective

_clusters

_d

_D

_dists

_k

_l

_m

_mbest

_mcurrent

_minDeviation

_x

_assignPoints

• _assignPoints(m: any, D: any): any[][]

  Parameters

  • m: any
  • D: any

  Returns any[][]

_findDimensions

• _findDimensions(m: any, L: any): any[][]

  Parameters

  • m: any
  • L: any

  Returns any[][]

_sample

• _sample(n: any, k: any): number[]

  Parameters

  • n: any
  • k: any

  Returns number[]

fit

• fit(): void

  Fit model.

  Returns void

init

• init(datas: number[][]): void

  Initialize model.

  Parameters

  • datas: number[][]

    Training data

  Returns void

outliers

• outliers(): boolean[]

  Returns a list of the data predicted as outliers or not.

  Returns boolean[]

  Predicted values

predict

• predict(): number[]

  Returns predicted categories.

  Returns number[]

  Predicted values