RandomSearchAutoML<T, TInput, TOutput>

Models & Types · AiDotNet.AutoML

AutoML implementation that uses random search over candidate model types and hyperparameters.

Inheritance: AutoMLModelBase<T, TInput, TOutput> → SupervisedAutoMLModelBase<T, TInput, TOutput> → BuiltInSupervisedAutoMLModelBase<T, TInput, TOutput> → RandomSearchAutoML<T, TInput, TOutput>

For Beginners

This AutoML strategy works like this:

  • Pick a model type at random (for example, Random Forest or Logistic Regression).
  • Pick a set of settings at random (for example, number of trees).
  • Train the model and score it on validation data.
  • Repeat and keep the best result.

If you are new to AutoML, random search is a good first choice because it is reliable and easy to reason about.

How It Works

Random search is a strong baseline for AutoML. It is simple, parallelizable, and often competitive with more complex search strategies for a given compute budget.

Example

var valTargets = new Vector<double>(new double[] { 0.0, 1.0, 0.0, 1.0 });
var valInputs = new Matrix<double>(new double[,] { { 1.0, 2.0 }, { 3.0, 4.0 }, { 5.0, 6.0 }, { 7.0, 8.0 } });
var trainTargets = new Vector<double>(new double[] { 0.0, 1.0, 0.0, 1.0 });
var trainInputs = new Matrix<double>(new double[,] { { 1.0, 2.0 }, { 3.0, 4.0 }, { 5.0, 6.0 }, { 7.0, 8.0 } });
var automl = new RandomSearchAutoML<double, Matrix<double>, Vector<double>>();
var bestModel = await automl.SearchAsync(
    trainInputs, trainTargets,
    valInputs, valTargets,
    
    timeLimit: TimeSpan.FromMinutes(5));

Constructors

Constructor Summary
RandomSearchAutoML(Random random)

Methods

Method Summary
Task<Dictionary<string, object>> SuggestNextTrialAsync()
Task<IFullModel<T, TInput, TOutput>> SearchAsync(TInput inputs, TOutput targets, TInput validationInputs, TOutput validationTargets, TimeSpan timeLimit, CancellationToken cancellationToken)