Based on the possibility of class output, machine learning classification can be categorized into binary classification, multiclass classification, and multilabel classification, as follows:

• Binary classification: This classifies observations into one of two possible classes. The example of spam email filtering we mentioned earlier is a typical use case of binary classification, which identifies email messages (input observations) as spam or not spam (output classes). Customer churn prediction is another frequently mentioned example, where the prediction system 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 click-through prediction for online ads—that is, whether or not an ad will be clicked, given users' cookie information and browsing history. Last, but not least, binary classification has also been employed in biomedical science, for example, in early cancer diagnosis, classifying patients into high or low risk groups based on MRI images. As demonstrated in the following example, binary classification tries to find a way to separate data from two classes:

• Multiclass classification: This classification is also called multinomial classification, and allows more than two possible classes, as opposed to only two in binary cases. Handwritten digit recognition is a common instance of classification and 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 from 0 to 9 in most countries) by which envelopes are automatically sorted. Handwritten digit recognition has become a hello world in the journey of studying machine learning, and the scanned document dataset constructed from the National Institute of Standards and Technology, called MNIST (short for Modified National Institute of Standards and Technology), is a benchmark dataset frequently used to test and evaluate multiclass classification models. The following screenshot shows the four samples taken from the MNIST dataset:

In the following example, the multiclass classification model tries to find segregation boundaries to separate data from the following three different classes:

• Multi-label classification: This 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 the 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 cat or dog in a binary case, or one type of fruit among oranges, apples, and bananas in a multiclass case. 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 problem, which is then handled by individual binary classifiers. Refer to the following diagram of restructuring a multi-label classification problem into multiple binary classification problems: