Class RankNet

constructor

• new RankNet(
      layer_sizes: number[],
      activations?: string | string[],
      rate?: number,
  ): RankNet

  Parameters

  • layer_sizes: number[]

    Sizes of all layers
  • Optionalactivations: string | string[]

    Activation names
  • Optionalrate: number

    Learning rate

  Returns RankNet

_a

_activations

_b

_layer_sizes

_optimizer

_rate

_w

_calc

• _calc(x: any): any[][]

  Parameters

  • x: any

  Returns any[][]

_init

• _init(sizes: any): void

  Parameters

  • sizes: any

  Returns void

fit

• fit(x1: number[][], x2: number[][], comp: (-1 | 0 | 1)[]): number

  Fit model.

  Parameters

  • x1: number[][]

    Training data 1
  • x2: number[][]

    Training data 2
  • comp: (-1 | 0 | 1)[]

    Sign of (data 1 rank - data 2 rank). If data 1 rank is bigger than data 2, corresponding value is 1.

  Returns number

  loss

predict

• predict(x: number[][]): number[]

  Returns predicted values.

  Parameters

  • x: number[][]

    Sample data

  Returns number[]

  Predicted values