[8] Understanding K-Neighborhood (KNN)

KNN KNN belongs to a typical 'Lazy Learner'. That is, instead of learning the discriminant function from the training data, we proceed with the learning by storing the training dataset in memory. Thanks to this, there is no cost in the learning process. Instead, the computational cost in the prediction phase is high. Memory-based classifiers have the advantage of being able to adapt immediately ..

2021.01.17