Class MarkovSwitching

constructor

• new MarkovSwitching(regime: number): MarkovSwitching

  Parameters

  • regime: number

    Number of regime

  Returns MarkovSwitching

_last_prob

_mu

_regime

_sigma

_lh

• _lh(x: any, mu: any, sigma: any): Matrix<T>

  Parameters

  • x: any
  • mu: any
  • sigma: any

  Returns Matrix<T>

_logL

• _logL(x: any, mu: any, sigma: any, prob: any): number

  Parameters

  • x: any
  • mu: any
  • sigma: any
  • prob: any

  Returns number

_mcmc

• _mcmc(x: any, eps: any, trial: any): (number[] | any[][] | Matrix<number>[])[]

  Parameters

  • x: any
  • eps: any
  • trial: any

  Returns (number[] | any[][] | Matrix<number>[])[]

_nextParam

• _nextParam(genProb: any, eps: any): any[]

  Parameters

  • genProb: any
  • eps: any

  Returns any[]

_prob

• _prob(x: any): Matrix<T>[]

  Parameters

  • x: any

  Returns Matrix<T>[]

_stationary_prob

• _stationary_prob(p: any): Matrix<number>

  Parameters

  • p: any

  Returns Matrix<number>

fit

• fit(datas: number[][], eps: number, trial: number): void

  Fit model.

  Parameters

  • datas: number[][]

    Training data
  • eps: number

    Parameter update range
  • trial: number

    Trial count

  Returns void

predict

• predict(datas: number[][]): number[]

  Returns anomaly degrees.

  Parameters

  • datas: number[][]

    Sample data

  Returns number[]

  Predicted values

probability

• probability(datas: number[][]): number[][]

  Returns probabilities.

  Parameters

  • datas: number[][]

    Sample data

  Returns number[][]

  Predicted values