Class KernelPCA

constructor

• new KernelPCA(
      kernel:
          | "gaussian"
          | "polynomial"
          | { name: "gaussian"; sigma?: number }
          | { n?: number; name: "polynomial" }
          | ((arg0: number[], arg1: number[]) => number),
      rd?: number,
  ): KernelPCA

  Parameters

  • kernel:
        | "gaussian"
        | "polynomial"
        | { name: "gaussian"; sigma?: number }
        | { n?: number; name: "polynomial" }
        | ((arg0: number[], arg1: number[]) => number)

    Kernel name
  • Optionalrd: number

    Reduced dimension

  Returns KernelPCA

_e

_kernel

_rd

_x

_gram

• _gram(x: any): Matrix<T>

  Parameters

  • x: any

  Returns Matrix<T>

fit

• fit(x: number[][]): void

  Fit model.

  Parameters

  • x: number[][]

    Training data

  Returns void

predict

• predict(x: number[][]): number[][]

  Returns reduced datas.

  Parameters

  • x: number[][]

    Sample data

  Returns number[][]

  Predicted values