PNAAggregator

Enums · AiDotNet.Enums

Aggregation function types for Principal Neighbourhood Aggregation (PNA).

For Beginners

These are different ways to combine information from neighbor nodes:

  • Mean: Average all neighbor features (balanced, smooth)
  • Max: Take the maximum value (emphasizes strong signals)
  • Min: Take the minimum value (emphasizes weak signals)
  • Sum: Add up all features (sensitive to number of neighbors)
  • StdDev: Standard deviation (captures variance in neighborhood)

Fields

Field Summary
PNAAggregator Max Max aggregation - takes maximum of neighbor features.
PNAAggregator Mean Mean aggregation - averages neighbor features.
PNAAggregator Min Min aggregation - takes minimum of neighbor features.
PNAAggregator StdDev Standard deviation aggregation - computes std of neighbor features.
PNAAggregator Sum Sum aggregation - sums neighbor features.
int value__