委员会方法（集成学习者）最佳模型
Committee Methods (ensemble learners) Best Model
我有以下问题，希望得到您的支持：
You are given the choice of two types of base classifiers (models) to
use in a (bagged) committee:
- Classifier A: which is known to have an (error) variance of 40, but has been found to produce uncorrelated models under bagging
- Classifier B: which is known to have an (error) variance of 30, but has been found to produce models with a pairwise covariance of 0.3
under bagging
Assuming Classifiers A and B have equal error bias and a committee of
size 10, which Classifier would you choose and why?
就个人而言，我会选择分类器 B，因为方差为 0.3 比不相关的模型更好。
我说的对吗？
你是对的。由于您指定的原因，B 更好
我有以下问题，希望得到您的支持：
You are given the choice of two types of base classifiers (models) to use in a (bagged) committee:
- Classifier A: which is known to have an (error) variance of 40, but has been found to produce uncorrelated models under bagging
- Classifier B: which is known to have an (error) variance of 30, but has been found to produce models with a pairwise covariance of 0.3 under bagging
Assuming Classifiers A and B have equal error bias and a committee of size 10, which Classifier would you choose and why?
就个人而言，我会选择分类器 B，因为方差为 0.3 比不相关的模型更好。
我说的对吗？
你是对的。由于您指定的原因，B 更好