Class MLPRegressor

constructor

• new MLPRegressor(
      hidden_sizes: number[],
      activation?:
          | "gaussian"
          | "identity"
          | "elu"
          | "leaky_relu"
          | "sigmoid"
          | "softplus"
          | "softsign"
          | "tanh"
          | "relu",
  ): MLPRegressor

  Parameters

  • hidden_sizes: number[]

    Sizes of hidden layers
  • Optionalactivation:
        | "gaussian"
        | "identity"
        | "elu"
        | "leaky_relu"
        | "sigmoid"
        | "softplus"
        | "softsign"
        | "tanh"
        | "relu"

    Activation name

  Returns MLPRegressor

_activations

_epoch

_hidden_sizes

_model

epoch

• get epoch(): number

  Epoch

  Returns number

_fitonce

• _fitonce(x: any, y: any, r: any): Matrix<number>

  Parameters

  • x: any
  • y: any
  • r: any

  Returns Matrix<number>

fit

• fit(
      train_x: number[][],
      train_y: number[][],
      iteration: number,
      rate?: number,
      batch?: number,
  ): number

  Fit model.

  Parameters

  • train_x: number[][]

    Training data
  • train_y: number[][]

    Target values
  • iteration: number

    Iteration count
  • Optionalrate: number

    Learning rate
  • Optionalbatch: number

    Batch size

  Returns number

  Loss value

predict

• predict(x: number[][]): number[][]

  Returns predicted values.

  Parameters

  • x: number[][]

    Sample data

  Returns number[][]

  Predicted values

toObject

• toObject(): LayerObject[]

  Returns object representation.

  Returns LayerObject[]

  Object represented this neuralnetwork