- Classification is a machine learning task that predicts a class label for each input example.
- Four common types are binary, multi-class, multi-label and imbalanced classification.
- Classification predicts a discrete class label; regression predicts a continuous quantity.
- Classification assigns objects to predefined classes, while clustering finds groups without predefined labels.
- Banks use classification to sort loan applicants into likely defaulters and non-defaulters.
What is Classification?
Classification is one of the primary uses of data science and machine learning. It is a systematic grouping of observations into categories. Examples could include when scientists categorize plants and animals into different taxonomies.
Examples of Classification
Below is a list of four types of classification and a brief example of each.
Binary Classification: Classification tasks that contain two class labels, such as:
- Email spam detection: spam or not
- Churn prediction: churn or not
Multi-Class Classification: multi-class classification refers to classification tasks that have more than two class labels, such as:
- Plant species classification
- Face classification
Multi-Label Classification: refers to classification tasks that have two or more class labels, where one or more class labels may be predicted for each example.
Imbalanced Classification: refers to classification tasks where the number of examples in each class is unequally distributed, such as:
- Medical diagnostic tests
- Fraud detection
Why is Classification Important?
Machine learning classification has many business applications. Take for example banking loans and the likelihood that someone will default on a loan. If you want to determine the chances of someone defaulting on a loan, you need to determine if that person belongs to one of the two classes with similar characteristics: the default class or the non-defaulter class. This classification method helps banks understand how likely the person is to default on the loan and helps banks adjust their risk assessment.
Classification vs Other Methodologies
Classification vs regression
Classification is the task of predicting a discrete class label. Regression is the task of predicting a continuous quantity.
Classification vs clustering
Classification uses predefined classes in which objects are assigned. Clustering identifies similarities between objects, which it groups according to those characteristics in common which also differentiate them from other groups of objects. These groups are known as clusters.
Classification vs prediction
Classification is the process of identifying the category or class label of the new observation to which it belongs. Prediction is the process of identifying the missing or unavailable numerical data for a new observation.
Classification in H2O AI
H2O Driverless AI, our AutoML, builds binary and multiclass classification models for uses such as probability of default, churn and fraud detection. Open-source H2O-3 reports classification metrics including AUC, AUCPR, F1, logloss and the Gini coefficient. H2O-3 classification metrics.