What is plurality classification in decision trees?

Question

I am new to the field of AI and am reading about decision trees. I am referring to the AIMA book which is pretty much the standard Intro to AI book recommended. In the chapter on decision trees, they discuss in the book a case wherein after the first attribute splits and there are no attributes left but both positive and negative examples have still not been separated, it means that these examples have exactly the same description... The solution to this case that they suggest is "to return the plurality classification of the remaining examples". I was wondering what that part in bold means? What does it mean to return the 'plurality classification' of a set of examples?

Answer 1

In decision trees when you have reached a leaf node but still do not have a clear idea about the class to assign it to, then you have to return plurality, which means consider all the examples of the leafs parent and see the most common class occurred in the dataset.

Plurality is just the generalization of the majority to more than 2 classes. It means to take the most frequent class in that leaf and return that as your prediction. So if we have 10 positive sentences and 5 negative sentences unseparated and there are 0 remaining attributes to split them, then the algorithm would just return positive (YES) as the value for these attributes.

Decision Tree learning is one of the predictive modeling approaches used in Data Mining and Machine Learning So if you wish to learn more about Decision Tree visit this Machine Learning Course.