|
aArtificial Neural Networks: An ANN is a network of simple processors which might have a small amount of memory space. These processors are connected by one-direction communications channels (these are 'connections' which carry numeric data. These single units work on their data stored in the memory and any inputs they receive through the connections. The network is principally a processing device based on the design and functions of a human brain. Most have cumulative learning ability and subsequently 'learn' from examples. They have a large potential for parallelism. ANNs can do all an ordinary digital computer can do and are particularly good at mapping problems which may have a few errors and have lots of data examples. Cellular Automata: Cellular Automata are systems, which use a three-dimensional grid for reference and information gathering using x, y and z co-ordinates. This table contains the laws and theories of the universe, which the computer uses to influence its actions, as these laws are always correct and uniform. These cells are cumulative which means that they remember previous processes and happenings to predict future occurrences. Dynamical systems use various inputs, which are processed to produce output information. The output completes a feed back loop, which the computer uses to process its next input information.
|