Options

# How to explain the binary2multiclass

Member Posts: 7 Contributor II
Hello good day! I am using a polynominal by binomial operator inside of it is the logistic regression now the result of the model is like this and I really dont know how to interpret or just explain the result:

I believe I don't need to upload the dataset. I know that Positive coefficient make the event more likely and negative coefficients make the event less likely. I am by the way using a multiclass here, there are 3 values in my independent variable Low Mastery, Very Low Mastery and Average Mastery which doesn't make it as binary value ,I really don't how to explain the 3 versus which is above my model. Does this mean there are 3 models?. The attribute there corresponds to the subject and the three values, Low Mastery, Very Low Mastery and Average Mastery are values of their National Achievement Test. What does these 3 events (Low Mastery vs. Very Low Mastery, Low Mastery vs. Average Mastery, Very Low Mastery vs. Average Mastery ) mean? . How does these 3 related to my model?

• Options
Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
This operator is creating separate models for each pairwise comparison of your multi-class label.  So if you have 3 classes, it does a model for 1 vs 2, 2 vs 3, and 1 vs 3.

Brian T.
Lindon Ventures
Data Science Consulting from Certified RapidMiner Experts
• Options
Member Posts: 7 Contributor II
@Telcontar120
Thank you so much for giving me the Idea of pairwise comparison for this multi class label. However, I got a difficulty of explaining for example the coefficient of each model. Does this also mean that for one model, Low Mastery vs Very Low mastery the negative coefficient means that as the independent variable increases, the dependent variable tends to decrease
• Options
Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
Yes, that is correct.  You will have to look at your coding to see which one is which, but that is the proper interpretation of the coefficients.

Brian T.
Lindon Ventures
Data Science Consulting from Certified RapidMiner Experts
• Options
Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
Yes, that is correct.  You will have to look at your coding to see which one is which, but that is the proper interpretation of the coefficients.

Brian T.
Lindon Ventures
Data Science Consulting from Certified RapidMiner Experts