Options

false and true in binominal attributes

soheil08soheil08 Member Posts: 2 Contributor I
edited August 2020 in Help

Hi
There are some nominal attributes in my data set that have just two value and in general they are binary attributes. For extracting FPGrowth and association rules I convert all of them to binominal attribute. For example attribute sex has two value "M" or "F" but my problem is I don't know which one is considered true and which one is considered false!
for example in association rules results I have [sex] --> [....] but I don't know what does it mean!

Help
Tagged:

Answers

  • Options
    JEdwardJEdward RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 578 Unicorn
    You have two options when converting to binominal. 

    1. (as you have done) turned them into True & False values by replacing.  You can specify in the operator which value represents true (for example Male=True). 
    2. Create two attributes named Sex_Male, Sex_Female and then they will each have opposite True & False values. 

    I would perhaps consider using option two for your task.  Unless it is market basket on selling a product primarily targetting women's clothing as then you can map the smallest class (Male) as true & use this to help filter out the reccomendations specifically for the 'Gift shoppers'. 
Sign In or Register to comment.