Constraint Processing
Fuzzy constraint networks for signal pattern recognition
Artificial Intelligence - Special issue: Fuzzy set and possibility theory-based methods in artificial intelligence
Hi-index | 0.00 |
The Multivariable Fuzzy Temporal Profile model enables human experts to project their knowledge of a signal pattern, described over a set of parameters, into a computable description. Here it is endowed with a hierarchical capability through the definition of algorithms that recognize a set of findings over a signal, and aggregate them into a set of abstraction levels. We also present a heuristics that takes advantage of the continuity properties of real signals to allow the matching process to meet real time requirements, even over high frequency signals.