Normally, classifiers can be categorized as the supervised learning. All
classifiers require a dataset which contains one attribute (mostly nominal) as a target
or a class with several distinct values. Then, classifiers learn patterns, rules or
characteristics in a given dataset that can categorize instances into different classes.
Methods to learn these patterns, rules or characteristics are different depending on
each algorithm. Users can use these patterns or rules given by algorithms to classify
instances with the unknown class in the same problem. Data mining models which
are widely used for classification problems are decision tree, naïve Bayes model,
support vector machine, neural network and k-nearest neighbor.