Support vector machines: a tool for pattern recognition and classification

Abstract : The power of computation and large memory of computers nowadays offer a great opportunity for information processing and storage. But information is not knowledge and one needs methods that permit to go from information to knowledge. Extracting automatically knowledge from storage data becomes then one of great challenge for the Information Technology (IT) industry. Pattern Recognition (PR) is the study of how machines can observe the environment, learn to distinguish pattern of interest from their background and make sound and reasonable decisions about the category of the pattern. The automatic recognition, classification, description, grouping of pattern is an important problem in engineering and sciences such as biology, psychology, medicine, marketing, computer vision, artificial intelligence, remote sensing, manufacturing, etc. Computer programs that help many professional in their daily work such as doctors diagnosing disease, policemen identifying suspects, engineers supervising manufacturing plants and energy production systems, etc. depend in some way on pattern recognition. One important field and goal of pattern recognition is classification: supervised or unsupervised also known as clustering. In this paper we present a mathematical tool named support vector machines (SVM) that permit to derive efficient algorithms of learning and classification.
Document type :
Journal articles
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02134723
Contributor : Open Archive Toulouse Archive Ouverte (oatao) <>
Submitted on : Monday, May 20, 2019 - 4:51:33 PM
Last modification on : Thursday, May 23, 2019 - 1:01:10 AM

File

Tchangani_23638.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02134723, version 1
  • OATAO : 23638

Collections

Citation

Ayeley Tchangani. Support vector machines: a tool for pattern recognition and classification. Studies in Informatics and Control, Informatics and Control Publications, 2005, 14 (2), pp.99-110. ⟨hal-02134723⟩

Share

Metrics

Record views

4

Files downloads

5