Understanding Decision Stump in Adaboost

Question

I am trying to understand how Adaboost works, and many of the tutorials online involve the use of a decision tree stump. For example, http://www.cc.gatech.edu/~thad/6601-gradAI-fall2013/boosting.pdf or http://www.fromdev.com/2012/04/creating-weak-learner-with-decision.html. However, I am confused on how the plot is actually generated. Say I have the following training set:

               Event 1     Event 2     Event 3   
Classifier 1      -1          1          1
Classifier 2      1           -1         1
Classifier 3      1           1          -1

How should I generate a decision stump for this?

Answer 1

I wonder if you what decision stump is.

Decision stump is simply a decision tree with only one node. It is usually considered as the weak classifier in boosting algorithms since due to its simplicity it will not overfit.

In your question it is not clear what is even 1 .. 3. Anyhow, the classifiers can be : decision stump1, decision stump2 and decision stump3.