Deep.Net


PCA

Namespace: Tensor
Parent Module: Decomposition

Principal Component Analysis (PCA)

Constructors

ConstructorDescription
new()
Signature: unit -> PCA

CompiledName: .ctor

Static members

Static memberDescription
Perform(data, nComps)
Signature: (data:Tensor<'T> * nComps:int64 option) -> Tensor<'T> * PCAInfo<'T>
Type parameters: 'T

Apply Principal Component Analysis (PCA) whitening. data must be of the form [sample, feature]. nComps optionally specifies how many components to keep. Returns a tensor of the form [sample, component].

Reverse(whitened, arg2)
Signature: (whitened:Tensor<'T> * PCAInfo<'T>) -> Tensor<'T>
Type parameters: 'T

Reverses PCA whitening. whitened must be of the form [sample, component].

Fork me on GitHub