Loss

AbstractLoss

Every loss struct is a subtype of AbstractLoss

MSE

Represents the Mean Squared Error : L(y,y^i)=∑(y-y^i)^2

BinaryCrossentropy

Represents the Binary crossentropy error function

CategoricalCrossentropy

Represents the Categorical crossentropy error function

compute_loss(lossF::MSE, predicted::Tensor, target::Union{Tensor,AbstractArray,Float64,Int64})
compute_loss(lossF::BinaryCrossentropy, predicted::Tensor, target::Union{Tensor,AbstractArray,Float64,Int64})
compute_loss(lossF::CategoricalCrossentropy, predicted::Tensor, target::Union{Tensor,AbstractArray,Float64,Int64})

Compute the loss according to the lossF given.