2. Kernels and pattern recognition
Kernel machines were developed to solve statistical learning problems. To characterize these problems, we will present three "case studies" in which kernel methods have been used (or can be used) successfully. In these examples, it is important to clearly formalize the problems and identify their specificities and difficulties, so as to understand how to implement kernel machines.
2.1 Three typical examples
The notion of learning is linked to the ability of a system to improve itself through new experiences. Among the various facets of this learning, we are interested in those linked to pattern recognition in the broadest sense: perception capacities, conceptualization faculties, the ability to predict the behavior of a phenomenon. Perception relates in particular...
The Ultimate Scientific and Technical Reference
This article is included in
Software technologies and System architectures
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
Kernels and pattern recognition
Tools
References
A few books to find out more
The Ultimate Scientific and Technical Reference
