FinancialAutoML<T>

Models & Types · AiDotNet.Finance.AutoML

AutoML implementation for finance models (forecasting and risk).

Inheritance: AutoMLModelBase<T, Tensor<T>, Tensor<T>> → SupervisedAutoMLModelBase<T, Tensor<T>, Tensor<T>> → FinancialAutoML<T>

For Beginners

This class is a “model picker” for finance tasks. It tries several finance models and chooses the one that scores best on your data.

How It Works

FinancialAutoML searches across a curated set of finance models while preserving the facade pattern. You provide the architecture and budget; AutoML selects the model.

Example

var trainInputs = Tensor<float>.CreateRandom(64, 8);
var trainTargets = Tensor<float>.CreateRandom(64, 1);
var valInputs = Tensor<float>.CreateRandom(16, 8);
var valTargets = Tensor<float>.CreateRandom(16, 1);
            
var automl = new FinancialAutoML<float>();
var bestModel = await automl.SearchAsync(
    trainInputs, trainTargets,
    valInputs, valTargets,
    timeLimit: TimeSpan.FromMinutes(30));

Constructors

Constructor Summary
FinancialAutoML(FinancialAutoMLOptions<T> options, Random random) Initializes a new FinancialAutoML instance.

Methods

Method Summary
Task<Dictionary<string, object>> SuggestNextTrialAsync() Suggests the next trial parameters.
Task<IFullModel<T, Tensor<T>, Tensor<T>>> SearchAsync(Tensor<T> inputs, Tensor<T> targets, Tensor<T> validationInputs, Tensor<T> validationTargets, TimeSpan timeLimit, CancellationToken cancellationToken) Runs the AutoML search loop.