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

blue · 2020年01月11日

问一道题:NO.PZ2015120204000034 [ CFA II ]

问题如下图:

选项:

这题讲义上也没说清楚,第一个观点不对吧,CART也需要设置regularization parameter,这也是超参数呀?A.

B.

C.

解释:

1 个答案
已采纳答案

星星_品职助教 · 2020年01月11日

同学你好,

CART是可以设置超参数的,但是这个不是CART算法可以使用的必须前提条件。只有在当建模者认为CART有可能产生过度拟合问题的时候,可以选择加入超参数来降低overfitting。而加入超参数其实也不是唯一的选项,也可以在建立完CART后进行“剪枝”(pruning)。

blue · 2020年01月12日

简单明了,重点突出,赞

  • 1

    回答
  • 6

    关注
  • 678

    浏览
相关问题

NO.PZ2015120204000034 问题如下 Paul suggests the following step whiwoulrepeateevery quarter.Step 3 For eaof the 20 fferent groups, we use labeleta to train a mol thwill prethe five stocks (in any given group) thare most likely to become acquisition targets in the next one year. Comparing two ML mols thcouluseto accomplish Step 3, whistatement(s) best scribe(s) the aantages of using Classification anRegression Trees (CART) insteof K-Nearest Neighbor (KNN)? Statement I For CART there is no requirement to specify initihyperparameter (like K).Statement II For CART there is no requirement to specify a similarity (or stance) measure.Statement III For CART the output provis a visuexplanation for the prection A.Statement I only B.Statement III only C.Statements I, II anIII C is correct. The aantages of using CART over KNN to classify companies into two categories (“not acquisition target” an“acquisition target”), inclu all of the following: For CART there are no requirements to specify initihyperparameter (like K) or a similarity (or stance) measure with KNN, anCART provis a visuexplanation for the prection (i.e., the feature variables antheir cut-off values eano).A is incorrect, because CART provis all of the aantages incatein Statements I, II anIII. B is incorrect, because CART provis all of the aantages incatein Statements I, II anIII. CART也需要设置超参数吧?regularization parameters每一个no中的single feature和cutoff value不也是一种similarity吗?

2023-10-27 11:56 1 · 回答

NO.PZ2015120204000034 有点不明白为什么A是对的

2021-03-09 03:50 2 · 回答

老师,我和前面的同学问题差不多,我看了一下材料上,K-means clustering的优势里也写了visualize the ta

2020-08-06 13:24 2 · 回答

Statement III only Statements I, II anIII C is correct. The aantages of using CART over KNN to classify companies into two categories (“not acquisition target” an“acquisition target”), inclu all of the following: For CART there are no requirements to specify initihyperparameter (like K) or a similarity (or stance) measure with KNN, anCART provis a visuexplanation for the prection (i.e., the feature variables antheir cut-off values eano). A is incorrect, because CART provis all of the aantages incatein Statements I, II anIII. B is incorrect, because CART provis all of the aantages incatein Statements I, II anIII. 第二个statement是什么意思没理解

2020-03-14 08:53 1 · 回答

老师讲过,KNN也会provi a visuprection

2020-01-23 22:16 1 · 回答