NODE
| Collection | International Encyclopedia of Systems and Cybernetics |
|---|---|
| Year | 2004 |
| Vol. (num.) | 2(2) |
| ID | ◀ 2291 ▶ |
| Object type | General information, Methodology or model |
A connected spot or element in a graph or in a network.
Atoms in a complex molecule, words in a language, people in an organization and generally any elements more or less selectively interconnected with others, become nodes, i.e. add positional values to their proper nature.
H.M. PAYNTER thus explains the properties of the nodes: “Most generally they might be unlabeled and therefore uniform and undistinguishable. However, in any directed graph, some of the branches incident on a given node will be directed toward it, others away. But it will always be observed that for unlimited richness of structure all graph systems must have at least one (type of) node with three or more branches, and that if the graph is directed, at least two branches must be directed toward this node. Moreover, as the American philosopher and logician C. S. PEIRCE discovered nearly a century ago, but one triadic node-type, with two in-branches and one out-branch is necessary for infinite variety of structure” (1968, p.76).
This node model does now appear as absolutely fundamental for the understanding of neural nets, which in turn are basic models for any interpretation of complex systems formed by numerous interacting elements (each of which can be considered a node). On the other hand, PEIRCE'S model, which is the most elemental summation device, anticipated, for instance, the Mac CULLOCH-PITTS neural transmission model.
Complexity of networks grows factorially in relation to the number of their nodes and the connexions between these.
See also
Node (Triadic)