GraphSAGEOptions

Options & Configuration · AiDotNet.NeuralNetworks.Options

Configuration options for GraphSAGENetwork.

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → GraphEncoderOptions → GraphSAGEOptions

For Beginners

GraphSAGE learns by sampling a node’s neighbours and aggregating them, which lets it handle graphs too large to hold in memory at once. The values here are the ones it ships with, so you can use the model without configuring anything.

Constructors

Constructor Summary
GraphSAGEOptions() Initializes a new instance carrying the model’s published defaults.
GraphSAGEOptions(GraphSAGEOptions other) Initializes a new instance by copying another instance.

Properties

Property Summary
SAGEAggregatorType AggregatorType { get; set; } Gets or sets how a node’s sampled neighbours are combined.
bool Normalize { get; set; } Gets or sets whether each layer’s output embeddings are L2-normalised.
double DropoutRate { get; set; } Gets or sets the dropout rate.

Methods

Method Summary
void Validate() Throws if a value this model requires has been left unset or is not positive.