Class SVM

constructor

• new SVM(
      kernel:
          | "gaussian"
          | "linear"
          | { d?: number; name: "gaussian" }
          | { name: "linear" }
          | ((arg0: number[], arg1: number[]) => number),
  ): SVM

  Parameters

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

    Kernel name

  Returns SVM

_a

_alldata

_b

_C

_eps

_err

_kernel

_n

_t

_tolerance

_x

_fitOnce

• _fitOnce(all?: boolean): number

  Parameters

  • Optionalall: boolean

  Returns number

_predict1

• _predict1(data: any): number

  Parameters

  • data: any

  Returns number

fit

• fit(): void

  Fit model.

  Returns void

init

• init(train_x: number[][], train_y: (-1 | 1)[]): void

  Initialize this model.

  Parameters

  • train_x: number[][]

    Training data
  • train_y: (-1 | 1)[]

    Target values

  Returns void

predict

• predict(data: number[][]): number[]

  Returns predicted values.

  Parameters

  • data: number[][]

    Sample data

  Returns number[]

  Predicted values