Clementine学习

用Stat和Barchat来做数据了解
filter来去除不需要的字段
sample抽样
type：修改数据类型(set,flat,range)，预测类型(i/o/both)
feature selection
Then drag trained FS model to replace Feature selection
Then change Sample to ”discard sample" to use the rest 30%

then create antother Type and link to the built CHAID model. use Analysis, Evaluation

Then use clemrun.exe +.pim +.par

export pmml
Helper applications: can import ssas ALGORIGHMS,

?use chaid to explroe variable relation?