For humans and machines, underlying patterns are there behind the noise.
The broad term, "pattern recognition" can refer to any perception of meaningful regularities in a complex or 'noisy' environment. It is a field within computer science dating back at least to 1973, and the publication of "Pattern Classification and Scene Analysis," by Richard O. Duda and Peter E. Hart.
Hierarchical Temporal Memory
The neocortex is the outer layer of the mammalian brain.
One of the methods sometimes used to program pattern recognition is called "hierarchical temporal memory," (HTM) based upon a modelling of some of the properties of the neocortex.
Diagnosis of Cancer
One central motivation in the development of pattern recognition (PR) software has been the prospect of more efficient computer-driven diagnosis of disease. In July 2004, for example, Georgia Tech Research News ran a story about new treatments for cancer. The story makes the point that growing cancer tumors "need new capillaries to feed them nutrients and oxygen. By using pattern recognition software to study abnormal capillary changes, researchers might be able to detect early signs of cancer."
Lowering Stillborn Rate
Research in 2010 at the University of Uster, in Northern Ireland, has applied pattern recognition to ultrasound images in the hope of analyzing an unborn baby's movements and determining whether it is at risk for a stillbirth.
Tags: pattern recognition, pattern recognition software, recognition software