Class BlurringMeanShift

constructor

• new BlurringMeanShift(
      h: number,
      threshold: number,
      kernel?:
          | "gaussian"
          | "rectangular"
          | "triangular"
          | "epanechnikov"
          | "biweight"
          | "triweight"
          | { name: "gaussian" }
          | { name: "rectangular" }
          | { name: "triangular" }
          | { name: "epanechnikov" }
          | { name: "biweight" }
          | { name: "triweight" }
          | ((arg0: number) => number),
  ): BlurringMeanShift

  Parameters

  • h: number

    Smoothing parameter for the kernel

  • threshold: number

    Distance threshold for determining same cluster

  • Optionalkernel:
        | "gaussian"
        | "rectangular"
        | "triangular"
        | "epanechnikov"
        | "biweight"
        | "triweight"
        | { name: "gaussian" }
        | { name: "rectangular" }
        | { name: "triangular" }
        | { name: "epanechnikov" }
        | { name: "biweight" }
        | { name: "triweight" }
        | ((arg0: number) => number)

    Kernel name

  Returns BlurringMeanShift

_c

_h

_kernel

_threshold

_x

categories

• get categories(): number[]

  Category list

  Returns number[]

size

• get size(): number

  Number of clusters

  Returns number

_distance

• _distance(a: any, b: any): number

  Parameters

  • a: any
  • b: any

  Returns number

fit

• fit(): boolean

  Fit model.

  Returns boolean

  true if any centroids has moved

init

• init(data: number[][]): void

  Initialize model.

  Parameters

  • data: number[][]

    Training data

  Returns void

predict

• predict(): number[]

  Returns predicted categories.

  Returns number[]

  Predicted values