The classification algorithm is a supervised learning technique that allows us to classify data. In this algorithm we search for a targeted variable which allows us to categorize the data set into categories.
In the below figure, we can see that it is an example of mail filtering into spam, not-spam.
Then there are types of classification, that are distributed in the following categories:
- Binomial Classification
- Multinomial Classification
- Binomial Classification:
In this classification there are two classes that are classified. We can see that in the following graph. In that graph there are two points that are below that area classified into one class and the remaining one into one class. That is how classification works for binomial classification.
- Multinomial Classification:
In this classification there are more than two classes that are being classified for further calculations. And all the dots represent different classes.
Then there are types of learners that the classifiers are:
- Lazy Learner: Lazy learner simply stores the training data and waits until a testing data appears. The classification is done using the most related data in the stored training data. They have more predicting time than eager learners. E.g., K-Nearest Neighbor.
- Eager Learner: They construct a classification model based on the given training data before getting data for predictions. It must be to commit to a single hypothesis that will work the entire space. E.g., Decision tree, Naïve Bayes.