DropoutLayer class

A Dropout layer for regularization.

Dropout is a technique used to prevent overfitting. During training, it randomly sets a fraction of input units to 0 with a frequency of rate at each step. This forces the network to learn more robust features.

IMPORTANT: This layer should only be active during training. During evaluation or inference, it should be disabled or bypassed, allowing all data to pass through unmodified.

This implementation uses "inverted dropout," where the outputs of the non-dropped units are scaled up by 1 / (1 - rate). This ensures that the expected output magnitude remains the same, and no changes are needed at test time.

  • Input: A Tensor<Vector> or Tensor<Matrix>.
  • Output: A tensor of the same shape as the input.

Example

SNetwork model = SNetwork([
  DenseLayer(128, activation: ReLU()),
  DropoutLayer(0.5), // Drops 50% of the inputs from the previous layer
  DenseLayer(10),
]);
Inheritance

Constructors

DropoutLayer(double rate)

Properties

hashCode → int
The hash code for this object.
no setterinherited
isTraining ↔ bool
getter/setter pair
name ↔ String
A user-friendly name for the layer (e.g., 'dense', 'lstm').
getter/setter pairoverride-getter
parameters → List<Tensor>
A list of all trainable tensors (weights and biases) in the layer.
no setteroverride
rate ↔ double
getter/setter pair
runtimeType → Type
A representation of the runtime type of the object.
no setterinherited

Methods

build(Tensor input) → void
Initializes the layer's parameters based on the shape of the first input.
inherited
call(Tensor input) → Tensor
The public, callable interface for the layer.
inherited
forward(Tensor input) → Tensor<Vector>
The core logic of the layer's transformation.
override
noSuchMethod(Invocation invocation) → dynamic
Invoked when a nonexistent method or property is accessed.
inherited
toString() → String
A string representation of this object.
inherited

Operators

operator ==(Object other) → bool
The equality operator.
inherited