Skip to content

[source]

Adaline#

Adaptive Linear Neuron is a single layer feed-forward neural network with a continuous linear output neuron suitable for regression tasks. Training is equivalent to solving L2 regularized linear regression (Ridge) online using Mini Batch Gradient Descent.

Interfaces: Estimator, Learner, Online, Ranks Features, Verbose, Persistable

Data Type Compatibility: Continuous

Parameters#

# Name Default Type Description
1 batchSize 128 int The number of training samples to process at a time.
2 optimizer Adam Optimizer The gradient descent optimizer used to update the network parameters.
3 alpha 1e-4 float The amount of L2 regularization applied to the weights of the output layer.
4 epochs 1000 int The maximum number of training epochs. i.e. the number of times to iterate over the entire training set before terminating.
5 minChange 1e-4 float The minimum change in the training loss necessary to continue training.
6 window 5 int The number of epochs without improvement in the training loss to wait before considering an early stop.
7 costFn LeastSquares RegressionLoss The function that computes the loss associated with an erroneous activation during training.

Example#

use Rubix\ML\Regressors\Adaline;
use Rubix\ML\NeuralNet\Optimizers\Adam;
use Rubix\ML\NeuralNet\CostFunctions\HuberLoss;

$estimator = new Adaline(256, new Adam(0.001), 1e-4, 500, 1e-6, 5, new HuberLoss(2.5));

Additional Methods#

Return an iterable progress table with the steps from the last training session:

public steps() : iterable

use Rubix\ML\Extractors\CSV;

$extractor = new CSV('progress.csv', true);

$extractor->export($estimator->steps());

Return the loss for each epoch from the last training session:

public losses() : float[]|null

Return the underlying neural network instance or null if untrained:

public network() : Network|null

References#


  1. B. Widrow. (1960). An Adaptive "Adaline" Neuron Using Chemical "Memistors". 


Last update: 2021-05-08