Module engram::nn

Neural networks.

Structs

• Dataset
• DenseLayer
• L1
  L1 regularization with the given lambda value. Also known as LASSO, used to encourage sparsity.
• L2
  L2 regularization with the given lambda value. Also known as Ridge, used to encourage small weights.
• Sequential

Enums

• Initializer
  Initialization methods for weight matrices.

Traits

• Layer
• Regularization