LightGbmBinaryTrainer.Options.WeightOfPositiveExamples Field

  • Reference

Definition

Namespace:
Microsoft.ML.Trainers.LightGbm
Assembly:
Microsoft.ML.LightGbm.dll
Package:
Microsoft.ML.LightGbm v3.0.1
Package:
Microsoft.ML.LightGbm v1.0.0
Package:
Microsoft.ML.LightGbm v1.1.0
Package:
Microsoft.ML.LightGbm v1.2.0
Package:
Microsoft.ML.LightGbm v1.3.1
Package:
Microsoft.ML.LightGbm v1.4.0
Package:
Microsoft.ML.LightGbm v1.5.5
Package:
Microsoft.ML.LightGbm v1.6.0
Package:
Microsoft.ML.LightGbm v1.7.0
Package:
Microsoft.ML.LightGbm v2.0.0

Controls the balance of positive and negative weights in LightGbmBinaryTrainer.

public double WeightOfPositiveExamples;
val mutable WeightOfPositiveExamples : double
Public WeightOfPositiveExamples As Double

Field Value

Double

This is useful for training on unbalanced data. A typical value to consider is sum(negative cases) / sum(positive cases).

Applies to