Artificial Intelligence - Special issue on knowledge representation
A framework for knowledge-based temporal abstraction
Artificial Intelligence
Fuzzy sets in approximate reasoning, Part 1: inference with possibility distributions
Fuzzy Sets and Systems
A Sufficient Condition for Backtrack-Free Search
Journal of the ACM (JACM)
Constraint Processing
Fuzzy constraint networks for signal pattern recognition
Artificial Intelligence - Special issue: Fuzzy set and possibility theory-based methods in artificial intelligence
Temporal scenario modelling and recognition based on possibilistic logic
Artificial Intelligence - Special issue: Fuzzy set and possibility theory-based methods in artificial intelligence
TRACE, a graphical tool for the acquisition and detection of signal patterns
Expert Systems with Applications: An International Journal
A cusum-based multilevel alerting method for physiological monitoring
IEEE Transactions on Information Technology in Biomedicine
Hi-index | 0.00 |
We present the Multivariable Fuzzy Temporal Profile model (MFTP), a formal model which allows us to represent signal patterns and identify their occurrences over the temporal evolution of a set of physical parameters. The pattern comprises a set of findings, each one of which may, in turn, be a pattern, so that its recognition is organized into a hierarchy of abstraction levels, and ultimately they are associated with the appearance of certain distinctive morphologies - profiles - over each parameter. The patterns definition is obtained directly from humans experts, either with the help of a formal language or with a visual tool developed for this purpose. The model is based, on the one hand, on Fuzzy Set Theory, which allows the vagueness and imprecision which are characteristic of human knowledge to be modelled; and on the other hand, on the formalism of Constraint Satisfaction Problems (CSP), in order to obtain a representation capable of explicitly capturing the hierarchy of abstraction levels into which the recognition task is organized. We supply algorithms for analyzing the consistency of the information defined by the MFTP, and for the recognition of patterns over signal recordings.