Loss {Microsoft.VisualBasic.MachineLearning.XGBoost.train} .NET clr documentation

Loss

Description

目标函数(损失)抽象。

梯度约定:grad = ∂L/∂pred,hess = ∂²L/∂pred², 叶值 w* = -G / (H + lambda)。

Declare

            
# namespace Microsoft.VisualBasic.MachineLearning.XGBoost.train
export class Loss {
   # 损失名称(用于序列化与参数校验)
   Name: string;
}

        

.NET clr type reference tree

this clr type has no other .net clr type reference.
[Package {$package} version {$version} Index]