Based on the possibility of class output, machine learning classification can be categorized into binary classification, multiclass classification, and multi-label classification.
Binary classification is the problem of classifying observations into one of the two possible classes. One frequently mentioned example is email spam filtering, which identifies email messages (input or observation) as spam or not spam (output or classes). Customer churn prediction is also a typical use of binary classification, where it takes in customer segment data and activity data from CRM systems and identifies which customers are likely to churn. Another application in the marketing and advertising industry is online ads click-through prediction-whether an ad will be clicked or not, given user's cookie information and browsing history.
Lastly, binary classification has also been employed in the biomedical field, to name one instance, the early cancer diagnosis classifying patients into high or low risk groups based on MRI images.
Multiclass classification, also called multinomial classification, allows more than two possible classes, as opposed to only two classes in binary cases. Handwritten digit recognition is a common instance and it has a long history of research and development since the early 1900s. A classification system, for example, learns to read and understand handwritten zip codes (digits 0 to 9 in most countries) by which envelopes are automatically sorted. And handwritten digit recognition has become a Hello World in the journey of learning machine learning, and the scanned document dataset constructed from the Modified National Institute of Standards and Technology called MNIST (whose samples are shown as follows) is frequently used to test and evaluate multiclass classification models.
MNIST hand-written digits recognition:
Multi-label classification is different from the first two types of classification where target classes are disjointed. Research attention to this field has been increasingly drawn by the nature of omnipresence of categories in modern applications. For example, a picture that captures a sea and sunset can simultaneously belong to both conceptual scenes, whereas it can only be an image of either a cat or dog in binary cases, or one fruit among orange, apple, and banana in multiclass cases. Similarly, adventure films are often combined with other genres, such as fantasy, science fiction, horror, and drama. Another typical application is protein function classification, as a protein may have more than one function-storage, antibody, support, transport, and so on. One approach to solve an n label classification problem is to transform it into a set of n binary classifications problems, which is then handled by individual binary classifiers respectively as shown in the following diagram:
