Connectionism

The connectionism is a problem-solving approach in cybernetics and deals with the behavior of networked systems based on combinations of artificial information processing units. Behavior is understood as a product of a large number of interacting components that mutually beeinflussen.Mit using artificial neural networks is the growing out of apparent chaos order system simulated. Applications of connectionism include neurophysiology, psychology, biology, linguistics, neuro computer science, exercise science and artificial intelligence research.

Problem solving with connectionist systems

Problem solving is independent of the respective fields of application always consists of the following steps:

  • Collect information
  • Form model
  • Create forecast
  • Check the result

The step of modeling is undoubtedly the most difficult. Expert systems, simulations and numerical calculations require detailed knowledge of the system to be studied. Your constructivist approach is based on the hypothesis that systems are symbolically written algorithmisierbar or completely by successive pre- Exempt disassemble into subsystems certain structure.

In a connectionist model tries the ( external ) behavior of the system as a whole to reproduce by the context a large number of relatively simple and often quite similar units, which are connected in a dense network with each other. These units operate locally and communicate with others only via signals via connections.

The development of a connectionist model system is carried out for selected examples of the system to be examined so that it Alike shows the same behavior as his role model. So for these cases there is an isomorphism of behavior, the connectionist model system responds to inputs with the same issues as its real role model. Since the system behavior is not algorithmisiert, but is not clear how the connectionist model system works internally, its results always arise from the interaction of all elements. Here, the connectionist model system need not necessarily be isomorphic to the investigation. According to Smolensky representation of knowledge takes place subsymbolisch.

Subsymbolic hypothesis

Deriving knowledge comes from the interaction of a large number of units. This interaction does not allow a precise description on a conceptual level, but must be directly realized by modeling processors. The conceptual model of a connectionist system is fundamental and independent of a specific implementation. In addition to the known artificial neural networks, particularly the sensitivity model by Frederic Vester as an implementation of a connectionist view is worth mentioning.

Advantages of connectionist architectures

The main advantages of systems with connectionist architecture (eg the human brain ) are:

484863
de