Weka中的Jrip输出

时间:2014-03-29 19:33:41

标签: weka

我正在对weka进行比较研究。使用一些评估标准,如;平均条件数。

avg number of condition = number of conditions in the antecedent part / 
                          number of all generated rules

但我得到了JRIP算法的以下输出:

(Wifes_education = 4) and (Number_of_children_ever_born >= 3) and (Wifes_age >= 34) and (Wifes_age <= 42) => Contraceptive_method_used=2 (141.0/56.0)
(Wifes_education = 4) and (Number_of_children_ever_born >= 3) and (Wifes_age >= 29) and (Wifes_age <= 48) and (Wifes_age >= 44) => Contraceptive_method_used=2 (53.0/24.0)
(Wifes_age <= 35) and (Number_of_children_ever_born >= 3) => Contraceptive_method_used=3 (377.0/171.0)
(Wifes_age <= 30) and (Number_of_children_ever_born >= 1) and (Wifes_age <= 22) => Contraceptive_method_used=3 (122.0/59.0)

**=> Contraceptive_method_used=1 (780.0/316.0)**
Number of Rules : 5

最后一条规则(粗体规则)怎么样?它没有先前的部分! 我怎样才能获得最后一条规则中的条件数?

1 个答案:

答案 0 :(得分:1)

最后一条规则(隐含)条件“捕获”其他规则未涵盖的所有内容。它的条件是“其他一切”。