Class LaplacianEigenmaps

constructor

• new LaplacianEigenmaps(
      rd: number,
      affinity?:
          | "rbf"
          | "knn"
          | { k?: number; name: "rbf"; sigma?: number }
          | { k?: number; name: "knn" },
      laplacian?: "unnormalized" | "normalized",
  ): LaplacianEigenmaps

  Parameters

  • rd: number

    Reduced dimension
  • Optionalaffinity:
        | "rbf"
        | "knn"
        | { k?: number; name: "rbf"; sigma?: number }
        | { k?: number; name: "knn" }

    Affinity type name
  • Optionallaplacian: "unnormalized" | "normalized"

    Normalized laplacian matrix or not

  Returns LaplacianEigenmaps

_affinity

_ev

_laplacian

_rd

predict

• predict(x: number[][]): number[][]

  Returns reduced datas.

  Parameters

  • x: number[][]

    Training data

  Returns number[][]

  Predicted values