Improving the Forecasting Capability of Fuzzy Inductive Reasoning
by Means of Dynamic Mask Allocation
Keywords
- Fuzzy Systems
- Inductive Reasoning
- Dynamic Model Allocation
Abstract
This paper deals with a new extension of the Fuzzy Inductive Reasoning (FIR)
methodology that makes use of the estimate of the prediction error generated by FIR
for automated dynamic model selection during the simulation run. FIR can choose among
several models available in the model library, and dynamically selects the model that
is currently most appropriate for the task of making inductively future predictions of
system behavior on the basis of observed earlier behavior.
Interested in reading the
full paper?
(8 pages, 237,474 bytes, pdf)
Interested in viewing the
Powerpoint presentation?
Homepage
Last modified: November 29, 2006 -- © François Cellier