Interface INonLinearRegression<T>

Namespace

AiDotNet.Interfaces

Assembly

AiDotNet.dll

Defines the functionality for non-linear regression models in the AiDotNet library.

public interface INonLinearRegression<T> : IRegression<T>, IFullModel<T, Matrix<T>, Vector<T>>, IModel<Matrix<T>, Vector<T>, ModelMetadata<T>>, IModelSerializer, ICheckpointableModel, IParameterizable<T, Matrix<T>, Vector<T>>, IFeatureAware, IFeatureImportance<T>, ICloneable<IFullModel<T, Matrix<T>, Vector<T>>>, IGradientComputable<T, Matrix<T>, Vector<T>>, IJitCompilable<T>

Type Parameters

T

The numeric data type used for calculations (e.g., float, double).

Inherited Members

IFullModel<T, Matrix<T>, Vector<T>>.DefaultLossFunction

IModel<Matrix<T>, Vector<T>, ModelMetadata<T>>.Train(Matrix<T>, Vector<T>)

IModel<Matrix<T>, Vector<T>, ModelMetadata<T>>.Predict(Matrix<T>)

IModel<Matrix<T>, Vector<T>, ModelMetadata<T>>.GetModelMetadata()

IModelSerializer.Serialize()

IModelSerializer.Deserialize(byte[])

IModelSerializer.SaveModel(string)

IModelSerializer.LoadModel(string)

ICheckpointableModel.SaveState(Stream)

ICheckpointableModel.LoadState(Stream)

IParameterizable<T, Matrix<T>, Vector<T>>.GetParameters()

IParameterizable<T, Matrix<T>, Vector<T>>.SetParameters(Vector<T>)

IParameterizable<T, Matrix<T>, Vector<T>>.ParameterCount

IParameterizable<T, Matrix<T>, Vector<T>>.WithParameters(Vector<T>)

IFeatureAware.GetActiveFeatureIndices()

IFeatureAware.SetActiveFeatureIndices(IEnumerable<int>)

IFeatureAware.IsFeatureUsed(int)

IFeatureImportance<T>.GetFeatureImportance()

ICloneable<IFullModel<T, Matrix<T>, Vector<T>>>.DeepCopy()

ICloneable<IFullModel<T, Matrix<T>, Vector<T>>>.Clone()

IGradientComputable<T, Matrix<T>, Vector<T>>.ComputeGradients(Matrix<T>, Vector<T>, ILossFunction<T>)

IGradientComputable<T, Matrix<T>, Vector<T>>.ApplyGradients(Vector<T>, T)

IJitCompilable<T>.ExportComputationGraph(List<ComputationNode<T>>)

IJitCompilable<T>.SupportsJitCompilation

Extension Methods

DistributedExtensions.AsDistributedForHighBandwidth<T, TInput, TOutput>(IFullModel<T, TInput, TOutput>, ICommunicationBackend<T>)

DistributedExtensions.AsDistributedForLowBandwidth<T, TInput, TOutput>(IFullModel<T, TInput, TOutput>, ICommunicationBackend<T>)

DistributedExtensions.AsDistributed<T, TInput, TOutput>(IFullModel<T, TInput, TOutput>, ICommunicationBackend<T>)

DistributedExtensions.AsDistributed<T, TInput, TOutput>(IFullModel<T, TInput, TOutput>, IShardingConfiguration<T>)

Remarks

This interface extends IFullModel to provide specialized capabilities for non-linear regression, which is used to model relationships that don't follow a straight line.

For Beginners: Non-linear regression helps you find patterns in data that follow curves instead of straight lines.

What is non-linear regression?

• Linear regression finds relationships that follow straight lines (like y = mx + b)
• Non-linear regression finds relationships that follow curves or more complex patterns
• These could be exponential curves, bell curves, sine waves, or other non-straight patterns

Real-world examples where non-linear patterns appear:

• Population growth (often follows exponential curves)
• Learning curves (progress is fast at first, then slows down)
• Seasonal sales data (follows cyclical patterns)
• Chemical reactions (may follow logarithmic or exponential patterns)
• Disease spread (often follows S-shaped logistic curves)

When to use non-linear regression:

• When plotting your data shows a clear curve rather than a straight line
• When you know from theory that the relationship should be non-linear
• When linear models give poor results or don't make sense for your problem

This interface provides methods to create, train, and use non-linear regression models through the functionality inherited from IFullModel.