Waist circumference: continuous var: How to replace the missing value? by min, max or average?

Noor_94_Noor_94_ Member Posts: 4 Newbie
Tagged:

Answers

  • ceaperezceaperez Member Posts: 351 Unicorn
    Hi @Noor_94_
    you need to use the "Replace Missing Values" for that. 

    Best, 

    Cesar
  • Noor_94_Noor_94_ Member Posts: 4 Newbie
    Thank you. Yes, I know it needs to use the "Replace Missing Values", but how can it replace them? Need to use min, max or average?
  • ceaperezceaperez Member Posts: 351 Unicorn
    Hi @Noor_94_

    It depends of your dataset, and the analysis type. For example if you haven't outliers and high dispersed data a good selection may be replace the missing using the mean. if you have outliers a recomendation is to use the median. 
    another options could be use like replace with mode. In any case, one of the most powerful things about Rapidminer is that you can test the impact of multiple data preprocessing decisions in your models very quickly.
    Test your decisions againts test dataset and meassure the results.

    Best, 
    Cesar
    Noor_94_
Sign In or Register to comment.