From the course: Data Science Foundations: Data Assessment for Predictive Modeling

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Techniques for working with the top predictors

Techniques for working with the top predictors

From the course: Data Science Foundations: Data Assessment for Predictive Modeling

Start my 1-month free trial

Techniques for working with the top predictors

“

- [Instructor] I'm in KNIME with an unmodified version of the census data set, and I've started a new workflow. What we're going to do now is walk through how to establish the relationships in your strong predictors to then discuss them with a subject matter expert. My favorite technique, is to grow the top branch of a decision tree. I'm going to take KNIME's Decision Tree Learner, and hook it up to the data, double click to configure it, and my class column is going to be income, we can keep it as that, but I'm going to go down here to force route split column, and what you want to do is grab one of the variables that you think has a strong relationship with this, and realize that at this point you would know that because you would have done by variate relationships of all the payers with the target variable. For instance, gender is going to have a strong relationship with income, and I can maximize that. Let's remember what…

Contents