3. Which of these is not a way to handle missing data? there are circumstances w

Question

3. Which of these is not a way to handle missing data?

there are circumstances where each of the other answers could be an acceptable way to handle missing data

let the classifier handle it in its own way (e.g. define "missing" as a separate value for the variable)

drop the instances with missing data

substitute the mean

multiple imputation

set the missing data to some specific value (e.g. 0)

drop the attribute with missing data

4. If we estimate a OneR (1R) model where some of the predictor variables are numeric, what problem might we see?

numeric predictors have to be converted into nominal values for OneR to be estimated

the decision tree will have only two nodes

one numeric variable will be "cut" into an unreasonably large number of pieces

Explanation / Answer

3. Following is not a way to handle missing data

set the missing data to some specific value

4. If we estimate a OneR (1R) model where some of the predictor variables are numeric, what problem might we see?

the decision tree will have only two nodes

Related Questions

Navigate

• Browse (All)
• Browse 3
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.