Constructors
constructor
- new SVM(c?, coef0?, degree?, gamma?, nu?, p?, kernelType?, classWeights?): SVM
Parameters
Optional
c: number
Optional
coef0: number
Optional
degree: number
Optional
gamma: number
Optional
nu: number
Optional
p: number
Optional
kernelType: number
Optional
classWeights: Mat
Returns SVM
- new SVM(params): SVM
Parameters
params: {
c?: number;
classWeights?: Mat;
coef0?: number;
degree?: number;
gamma?: number;
kernelType?: number;
nu?: number;
p?: number;
}
Optional
c?: number
Optional
classWeights?: Mat
Optional
coef0?: number
Optional
degree?: number
Optional
gamma?: number
Optional
kernelType?: number
Optional
nu?: number
Optional
p?: number
Returns SVM
Properties
Readonly
c
c: number
Readonly
classWeights
Readonly
coef0
coef0: number
Readonly
degree
degree: number
Readonly
gamma
gamma: number
Readonly
isTrained
isTrained: boolean
Readonly
kernelType
kernelType: number
Readonly
nu
nu: number
Readonly
p
p: number
Readonly
varCount
varCount: number
Methods
calcError
- calcError(trainData, test): {
error: number;
responses: Mat;
} Returns {
error: number;
responses: Mat;
}
error: number
responses: Mat
getDecisionFunction
- getDecisionFunction(i): {
alpha: Mat;
rho: number;
svidx: Mat;
} Returns {
alpha: Mat;
rho: number;
svidx: Mat;
}
alpha: Mat
rho: number
svidx: Mat
getSupportVectors
- getSupportVectors(): Mat
Returns Mat
load
- load(file): void
Returns void
predict
- predict(sample, flags?): number
Parameters
sample: number[]
Optional
flags: number
Returns number
- predict(samples, flags?): number[]
Parameters
samples: Mat
Optional
flags: number
Returns number[]
save
- save(file): void
Returns void
setParams
- setParams(c?, coef0?, degree?, gamma?, nu?, p?, kernelType?, classWeights?): void
Parameters
Optional
c: number
Optional
coef0: number
Optional
degree: number
Optional
gamma: number
Optional
nu: number
Optional
p: number
Optional
kernelType: number
Optional
classWeights: Mat
Returns void
- setParams(args): void
Parameters
args: {
c?: number;
classWeights?: Mat;
coef0?: number;
degree?: number;
gamma?: number;
kernelType?: number;
nu?: number;
p?: number;
}
Optional
c?: number
Optional
classWeights?: Mat
Optional
coef0?: number
Optional
degree?: number
Optional
gamma?: number
Optional
kernelType?: number
Optional
nu?: number
Optional
p?: number
Returns void
train
- train(trainData, flags?): boolean
Returns boolean
- train(samples, layout, responses): boolean
Parameters
samples: Mat
layout: number
responses: Mat
Returns boolean
trainAsync
- trainAsync(trainData, flags?): Promise<boolean>
Returns Promise<boolean>
- trainAsync(samples, layout, responses): Promise<boolean>
Parameters
samples: Mat
layout: number
responses: Mat
Returns Promise<boolean>
trainAuto
- trainAuto(trainData, kFold?, cGrid?, gammaGrid?, pGrid?, nuGrid?, coeffGrid?, degreeGrid?, balanced?): Mat
Parameters
Optional
kFold: number
Optional
gammaGrid: ParamGrid
Optional
coeffGrid: ParamGrid
Optional
degreeGrid: ParamGrid
Optional
balanced: boolean
Returns Mat
trainAutoAsync
- trainAutoAsync(trainData, kFold?, cGrid?, gammaGrid?, pGrid?, nuGrid?, coeffGrid?, degreeGrid?, balanced?): Promise<Mat>
Parameters
Optional
kFold: number
Optional
gammaGrid: ParamGrid
Optional
coeffGrid: ParamGrid
Optional
degreeGrid: ParamGrid
Optional
balanced: boolean
Returns Promise<Mat>