Class CRF

Conditional random fields

Index

Constructors

Properties

_e _phi _w _x_cand _xn _y_cand _yn

Methods

_alpha _beta _p _psi _z fit predict probability

Constructors

constructor

new CRF(): CRF
Returns CRF

Properties

_e

_e: number

_phi

_phi: (x: any, y1: any, y0: any) => any[]

_w

_w: number[]

_x_cand

_x_cand: any[]

_xn

_xn: number

_y_cand

_y_cand: any[]

_yn

_yn: number

Methods

_alpha

_alpha(x: any, scaled?: boolean): any[][]
Parameters
- x: any
- Optionalscaled: boolean
Returns any[][]
- Defined in lib/model/crf.d.ts:13

_beta

_beta(x: any, scaled?: boolean): any[][]
Parameters
- x: any
- Optionalscaled: boolean
Returns any[][]
- Defined in lib/model/crf.d.ts:14

_p

_p(x: any, scaled?: boolean): any[][]
Parameters
- x: any
- Optionalscaled: boolean
Returns any[][]
- Defined in lib/model/crf.d.ts:16

_psi

_psi(x: any, y1: any, y0: any): number
Parameters
- x: any
- y1: any
- y0: any
Returns number
- Defined in lib/model/crf.d.ts:12

_z

_z(x: any): any
Parameters
- x: any
Returns any
- Defined in lib/model/crf.d.ts:15

fit

fit(x: any[][], y: any[][]): void
Fit model.
Parameters
- x: any[][]
  Training data
- y: any[][]
  Target values
Returns void
- Defined in lib/model/crf.d.ts:22

predict

predict(x: any[][]): any[][]
Returns predicted labels.
Parameters
- x: any[][]
  Sample data
Returns any[][]
Predicted values
- Defined in lib/model/crf.d.ts:35

probability

probability(x: any[], y: any[]): number
Returns probability P(y|x).
Parameters
- x: any[]
  Sample data
- y: any[]
  Target values
Returns number
Predicted values
- Defined in lib/model/crf.d.ts:29