开发者:上海品职教育科技有限公司 隐私政策详情

应用版本:4.2.11(IOS)|3.2.5(安卓)APP下载

Marina_0122 · 2020年01月29日

问一道题:NO.PZ201512020300000507

* 问题详情,请 查看题干

问题如下:

Assuming a Classification and Regression Tree (CART) model is used to accomplish Step 3, which of the following is most likely to result in model overfitting?

选项:

A.

Using the k-fold cross validation method

B.

Including an overfitting penalty (i.e., regularization term).

C.

Using a fitting curve to select a model with low bias error and high variance error.

解释:

C is correct. A fitting curve shows the trade-off between bias error and variance error for various potential models. A model with low bias error and high variance error is, by definition, overfitted.
A is incorrect, because there are two common methods to reduce over
fitting, one of which is proper data sampling and cross-validation. K-fold cross validation is such a method for estimating out-of-sample error directly by determining the error in validation samples.
B is incorrect, because there are two common methods to reduce over
fitting, one of which is preventing the algorithm from getting too complex during selection and training, which requires estimating an overfitting penalty.

为什么不选B 不太懂题目的意思


1 个答案
已采纳答案

星星_品职助教 · 2020年02月03日

同学你好,

题干问哪种方式可能会导致overfitting。增加一个penalty term可以降低变量也就是features的数量,是可以降低overfitting的。

  • 1

    回答
  • 0

    关注
  • 697

    浏览
相关问题

NO.PZ201512020300000507 问题如下 Assuming a Classification anRegression Tree (CART) mol is useto accomplish Step 3, whiof the following is most likely to result in mol overfitting? A.Using the k-folcross valition metho B.Inclung overfitting penalty (i.e., regularization term). C.Using a fitting curve to selea mol with low bierror anhigh varianerror. C is correct. A fitting curve shows the tra-off between bierror anvarianerror for various potentimols. A mol with low bierror anhigh varianerror is, finition, overfitteA is incorrect, because there are two common metho to reoverfitting, one of whiis proper ta sampling ancross-valition. K-folcross valition is sua methofor estimating out-of-sample error rectly termining the error in valition samples.B is incorrect, because there are two common metho to reoverfitting, one of whiis preventing the algorithm from getting too complex ring selection antraining, whirequires estimating overfitting penalty. 老师,这个K-fol验不是为了解决数据量不够或说unfitting的问题么?它不会造成over fitting?

2023-10-01 19:01 1 · 回答

Inclung overfitting penalty (i.e., regularization term). Using a fitting curve to selea mol with low bierror anhigh varianerror. C is correct. A fitting curve shows the tra-off between bierror anvarianerror for various potentimols. A mol with low bierror anhigh varianerror is, finition, overfitte A is incorrect, because there are two common metho to reoverfitting, one of whiis proper ta sampling ancross-valition. K-folcross valition is sua methofor estimating out-of-sample error rectly termining the error in valition samples. B is incorrect, because there are two common metho to reoverfitting, one of whiis preventing the algorithm from getting too complex ring selection antraining, whirequires estimating overfitting penalty. 题干是说在Cart模型下,哪种会导致overfitting。但是中比如B增加惩罚项,这个从性质上来说,如何能和CART模型一起用呢,感觉和题干没办法融合

2021-01-22 16:52 1 · 回答

请问 K-folcross valition 在讲义里哪里有提到?谢谢

2020-05-25 18:03 1 · 回答

能一下这题吗?不是很理解,谢谢

2020-02-01 11:00 1 · 回答