Class MatchingNetworksOptions<T, TInput, TOutput>

Namespace

AiDotNet.MetaLearning.Options

Assembly

AiDotNet.dll

Configuration options for Matching Networks algorithm.

public class MatchingNetworksOptions<T, TInput, TOutput> : IMetaLearnerOptions<T>

Type Parameters

T

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

TInput

The input data type (e.g., Matrix<T>, Tensor<T>).

TOutput

The output data type (e.g., Vector<T>, Tensor<T>).

Inheritance

object

MatchingNetworksOptions<T, TInput, TOutput>

Implements

IMetaLearnerOptions<T>

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

Matching Networks use attention mechanisms over the support set to classify query examples. It computes a weighted sum of support labels where weights are determined by an attention function that measures similarity between examples.

For Beginners: Matching Networks learn to pay attention to similar examples:

1. Encode all examples (support and query) with a shared encoder
2. For each query, compute attention weights with all support examples
3. Use similarity (cosine, dot product, etc.) for weights
4. Predict weighted sum of support labels (soft nearest neighbor)

Unlike ProtoNets which uses class prototypes, Matching Networks consider every support example individually.

Constructors

MatchingNetworksOptions(IFullModel<T, TInput, TOutput>)

Initializes a new instance of the MatchingNetworksOptions class with the required meta-model.

public MatchingNetworksOptions(IFullModel<T, TInput, TOutput> metaModel)

Parameters

metaModel IFullModel<T, TInput, TOutput>

The encoder to be trained (required).

Examples

// Create Matching Networks with minimal configuration
var options = new MatchingNetworksOptions<double, Tensor<double>, Tensor<double>>(myEncoder);
var matchingNets = new MatchingNetworksAlgorithm<double, Tensor<double>, Tensor<double>>(options);

// Create with custom attention function
var options = new MatchingNetworksOptions<double, Tensor<double>, Tensor<double>>(myEncoder)
{
    AttentionFunction = MatchingNetworksAttentionFunction.DotProduct,
    UseBidirectionalEncoding = true
};

Exceptions

ArgumentNullException

Thrown when metaModel is null.

Properties

AdaptationSteps

Gets or sets the number of adaptation steps.

public int AdaptationSteps { get; set; }

Property Value

int

Default is 1 (Matching Networks uses attention-based adaptation).

AttentionFunction

Gets or sets the attention function for computing similarity.

public MatchingNetworksAttentionFunction AttentionFunction { get; set; }

Property Value

MatchingNetworksAttentionFunction

Default is Cosine.

Remarks

For Beginners: - Cosine: Measures angle between vectors (most common) - DotProduct: Simple inner product - Euclidean: Uses distance (inverted so closer = higher) - Learned: Uses a neural network to compute similarity

CheckpointFrequency

Gets or sets how often to save checkpoints.

public int CheckpointFrequency { get; set; }

Property Value

int

DataLoader

Gets or sets the episodic data loader for sampling tasks. Default: null (tasks must be provided manually to MetaTrain).

public IEpisodicDataLoader<T, TInput, TOutput>? DataLoader { get; set; }

Property Value

IEpisodicDataLoader<T, TInput, TOutput>

EnableCheckpointing

Gets or sets whether to save model checkpoints.

public bool EnableCheckpointing { get; set; }

Property Value

bool

EvaluationFrequency

Gets or sets how often to evaluate the meta-learner.

public int EvaluationFrequency { get; set; }

Property Value

int

EvaluationTasks

Gets or sets the number of tasks to use for evaluation.

public int EvaluationTasks { get; set; }

Property Value

int

GradientClipThreshold

Gets or sets the maximum gradient norm for gradient clipping.

public double? GradientClipThreshold { get; set; }

Property Value

double?

Default is 10.0.

InnerLearningRate

Gets or sets the learning rate for the inner loop (not used in Matching Networks).

public double InnerLearningRate { get; set; }

Property Value

double

Default is 0.01.

L2Regularization

Gets or sets the L2 regularization strength.

public double L2Regularization { get; set; }

Property Value

double

Default is 0.0 (no regularization).

LossFunction

Gets or sets the loss function for training. Default: null (uses cross-entropy loss internally).

public ILossFunction<T>? LossFunction { get; set; }

Property Value

ILossFunction<T>

MetaBatchSize

Gets or sets the number of tasks to sample per meta-training iteration.

public int MetaBatchSize { get; set; }

Property Value

int

Default is 4.

MetaModel

Gets or sets the meta-model (encoder) to be trained. This is the only required property.

public IFullModel<T, TInput, TOutput> MetaModel { get; set; }

Property Value

IFullModel<T, TInput, TOutput>

Remarks

The meta-model should be a feature encoder that maps inputs to an embedding space where similar examples are close together.

For Beginners: This is typically a CNN for images or an MLP for tabular data. The encoder learns to produce embeddings where attention weights are meaningful.

MetaOptimizer

Gets or sets the optimizer for encoder updates. Default: null (uses built-in Adam optimizer with OuterLearningRate).

public IGradientBasedOptimizer<T, TInput, TOutput>? MetaOptimizer { get; set; }

Property Value

IGradientBasedOptimizer<T, TInput, TOutput>

NumClasses

Gets or sets the number of output classes.

public int NumClasses { get; set; }

Property Value

int

Default is 5.

NumMetaIterations

Gets or sets the total number of meta-training iterations.

public int NumMetaIterations { get; set; }

Property Value

int

Default is 1000.

OuterLearningRate

Gets or sets the learning rate for the outer loop (encoder training).

public double OuterLearningRate { get; set; }

Property Value

double

Default is 0.001.

RandomSeed

Gets or sets the random seed for reproducibility.

public int? RandomSeed { get; set; }

Property Value

int?

Temperature

Gets or sets the temperature for softmax attention.

public double Temperature { get; set; }

Property Value

double

Default is 1.0.

Remarks

Lower temperature makes attention weights more peaked (focused on one example). Higher temperature makes attention weights more uniform.

UseBidirectionalEncoding

Gets or sets whether to use bidirectional encoding.

public bool UseBidirectionalEncoding { get; set; }

Property Value

bool

Default is false.

Remarks

When enabled, uses bidirectional LSTM to encode sequences, allowing information to flow in both directions.

UseFirstOrder

Gets or sets whether to use first-order approximation.

public bool UseFirstOrder { get; set; }

Property Value

bool

Default is true since Matching Networks doesn't use gradient-based inner loop.

UseFullContextEmbedding

Gets or sets whether to use full context embedding.

public bool UseFullContextEmbedding { get; set; }

Property Value

bool

Default is false.

Remarks

When enabled, each example's embedding considers all other examples in the episode through attention mechanisms.

Methods

Clone()

Creates a deep copy of the Matching Networks options.

public IMetaLearnerOptions<T> Clone()

Returns

IMetaLearnerOptions<T>

A new MatchingNetworksOptions instance with the same configuration.

IsValid()

Validates that all Matching Networks configuration options are properly set.

public bool IsValid()

Returns

bool

True if the configuration is valid; otherwise, false.