In the previous sections, we discussed ensemble methods. What we didn't mention earlier was how the results are aggregated across the individual models prepared by the ensemble techniques. The concept that is used for this is called plurality, which is nothing but voting. The higher the vote a class gets, the higher the chances of it being the final class. Imagine if we had three models prepared during ensemble techniques and 10 possible classes (think of them as digits from 0 to 9). Each model would choose one class based on the highest probability it obtained. Finally, the class with the maximum number of votes would be selected. This is the concept of plurality. In practice, plurality tries to bring benefit to both k-NN and Naive Bayes classification algorithms.

In short, plurality voting is a method where the class with the most votes wins. It is one form of majority voting. 

So, every classifier gives exactly one vote. The class with the most votes is the prediction of the ensemble.