Class GraphWiseDominantLayerConfig

java.lang.Object
oracle.pgx.config.mllib.GraphWiseDominantLayerConfig
All Implemented Interfaces:
GraphWiseEmbeddingConfig
public class GraphWiseDominantLayerConfig extends Object implements GraphWiseEmbeddingConfig
Configuration class for Dominant layers. See GraphWiseDominantLayerConfigBuilder for a description of the hyperparameters.
Since:
23.2

  • Field Details

    • DEFAULT_ALPHA

      public static final double DEFAULT_ALPHA
      Default alpha parameter as 0.5: equal weight for feature and structure reconstruction losses
      See Also:

    • DEFAULT_PREDICTION_LAYER_CONFIGS

      public static final GraphWisePredictionLayerConfig[] DEFAULT_PREDICTION_LAYER_CONFIGS
      by default no prediction layer config

  • Constructor Details

    • GraphWiseDominantLayerConfig

      public GraphWiseDominantLayerConfig()

  • Method Details

    • getEmbeddingType

      public GraphWiseEmbeddingConfig.EmbeddingType getEmbeddingType()
      Specified by:
      getEmbeddingType in interface GraphWiseEmbeddingConfig

    • getAlpha

      public double getAlpha()
      Returns:
      configured weight for feature reconstruction loss.

    • setAlpha

      public void setAlpha(double alpha)
      Sets the weight for feature reconstruction loss.
      Parameters:
      alpha - the feature reconstruction weight. Must be between 0 and 1

    • getdecoderLayerConfigs

      public GraphWisePredictionLayerConfig[] getdecoderLayerConfigs()
      Gets the configuration objects for the decoder layers
      Returns:
      configurations
      Since:
      23.2

    • setDecoderLayerConfigs

      public final void setDecoderLayerConfigs(GraphWisePredictionLayerConfig... decoderLayerConfigs)
      Sets the configuration objects for the decoder layers
      Parameters:
      decoderLayerConfigs - the configuration objects
      Since:
      23.2