Class SILK

constructor

• new SILK(
      eta?: number,
      lambda?: number,
      c?: number,
      w?: number,
      kernel?:
          | "gaussian"
          | "polynomial"
          | { name: "gaussian"; s?: number }
          | { d?: number; name: "polynomial" }
          | ((arg0: number[], arg1: number[]) => number),
      loss?: "hinge" | "square" | "logistic" | "graph",
  ): SILK

  Parameters

  • Optionaleta: number

    Learning rate
  • Optionallambda: number

    Regularization constant
  • Optionalc: number

    Penalty imposed on point prediction violations.
  • Optionalw: number

    Buffer size
  • Optionalkernel:
        | "gaussian"
        | "polynomial"
        | { name: "gaussian"; s?: number }
        | { d?: number; name: "polynomial" }
        | ((arg0: number[], arg1: number[]) => number)

    Kernel name
  • Optionalloss: "hinge" | "square" | "logistic" | "graph"

    Loss type name

  Returns SILK

_w