GM-RKB: Classification Function

A classification function is a discrete value output-function whose function range is a categorical set (with categories).

AKA: Classifier, Categorical Function, Category-Valued Function, Class-Output Function, Categorical-Output Function, Classification Structure, Classification Scheme.
Context:
- Function Output: a Class Member.
- Function Range: a Categorical Set.
- It can be:
  - a Binary Classification Function.
  - a Multi-Valued Classification Function.
- It can be:
  - a Probabilistic Classification Function, such as on that predicts the Outcome of a Random Experiment with a Categorical Sample Space.
  - a Deterministic Classification Function, such as a Deductive Logic Formula.
- It can range from:
  - being a Simple Classification Function
  - to being a Complex Classification Function.
- It can be:
  - a Predictive Classifier, with a Probability Value.
  - a Heuristic Classifier.
  - a Categorical Random Variable.
- It can be:
  - a Generative Classification Function.
  - a Discriminative Classification Function.
- It can a:
  - Deduced Classification Function.
  - Induced Classification Function, (that was induced by a Classification Function Learning Task).
  - Heuristic Classification Function.
- It can be, based on the Input Structure,
  - a Vector Classifier.
  - a Tuple Classifier.
  - a Sequence Classifier, such as a Text Classification Function.
  - a Sequence Tagger, such as a POS Tagging Function.
  - a Graph Classifier.
- It can be, based on the Structure,
- It can be required by a Classification Task
- It can be used by a Classification System (e.g. a Mechanized Classifier or a Human Classifier).
Example(s):
- a Vector Classifier,
  - f(1.1,2.3,3.9) => +1.
- a Tuple Classifier,
  - f(1.1,Yellow,3.9) => Red.
- a Complex Input Classification Function, such as a Sequence Classifier (such as a Text Classification Function).
- a Complex Output Classification Function, such as a Sequence Tagger (such as a POS Tagging Function).
Counter-Example(s):
- f(1.5) => 5.1, a Real-Valued Function.
- f(1.1,Yellow,3.9) => 0.37, , a Real-Valued Function.
- f(1.5) => 1, an Integer-Valued Function.
- f(1.5) => Large, an Ordinal-Valued Function.
- f(1.5) => (3.1,2.3), a Vector-Valued Function.
See: Predictive Classifier, Concept Hierarchy.

References

2008

(Dextre Clarke & al, 2008) => Stella Dextre Clarke, Alan Gilchrist, Ron Davies and Leonard Will. (2008). "Glossary of Terms Relating to Thesauri and Other Forms of Structured Vocabulary for Information Retrieval." Willpower Information
- classification scheme
  - schedule of concepts, arranged by classification

2007

http://gondolin.rutgers.edu/MIC/text/how/catalog_glossary.htm
- A scheme, usually consisting of numbers or alphanumericor other notation that categorizes or subdivides a subject area or collection of materials. Most classification schemes were originally intended to organize physical items on the shelf. The result was a unique shelving location (call number) for each item that facilitated browsing of material by subject or author.

2005

(Woodley, 2005b) => Mary S. Woodley, Gail Clement, and Pete Winn. (2005). "DCMI Glossary." Dublin Core Metadata Initiative.
- classification
  - A logical scheme for arrangement of knowledge, usually by subject. Classification schema are alpha and/or numeric; for example, Library of Congress Classification, Dewey Classification, Universal Decimal Classification.

1998

(Kohavi & Provost, 1998) => Ron Kohavi, and Foster Provost. (1998). "Glossary of Terms." In: Machine Leanring 30(2-3).
- Classifier: A mapping from unlabeled instances to (discrete) classes. Classifiers have a form (e.g., decision tree) plus an interpretation procedure (including how to handle unknowns, etc.). Some classifiers also provide probability estimates (scores), which can be thresholded to yield a discrete class decision thereby taking into account a utility function.

1995

(Kohavi, 1995) => Ron Kohavi. (1995). "A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection." In: Proceedings of the Fourteenth International Joint Conference on Artificial Intelligence (IJCAI 1995).
- A classifier is a function that maps an unlabelled instance to a label using internal data structures. An inducer or an induction algorithm builds a classifier from a given dataset. CART and C 4 5 (Brennan, Friedman Olshen &. Stone 1984, Quinlan 1993) are decision tree inducers that build decision tree classifiers. In this paper we are not interested in the specific method for inducing classifiers, but assume access to a dataset and an inducer of interest.