Applying Fuzzy-Based Inductive Reasoning to Analyze Qualitatively the Dynamic Behavior of an Ecological System

Abstract

In the past decade, a variety of methodologies for representing and evaluating knowledge in qualitative terms have been developed, particularly within the field of Artificial Intelligence. Qualitative reasoning methodologies represent a valid alternative to quantitative modeling approaches when dealing with systems for which only imprecise or incomplete knowledge is available, as this is often the case with ecological systems. As most of the proposed qualitative reasoning methodologies have not yet been applied to anything but toy examples, it remains a challenge to apply those methodologies to real-world applications. In Biosphere 2, a closed ecological system, the level of O2 has dropped and the CO2 level has risen continuously during its first continuous closure in the years 1991 to 1993. The mechanisms of atmospheric carbon cycles have been studied in various research efforts in the past, and general rules describing them have been formulated. However, the specific situation within Biosphere 2, a closed ecosystem, might challenge the validity of these rules. The structure of the carbon cycle in Biosphere 2 is not yet well understood, although abundant data exist on some of its important fluxes and pools. Whereas deductive, quantitative as well as qualitative, megies need knowledge about the structure of the system under study to derive models capturing the behavioral patterns of the system, the fuzzy-based inductive reasoning methodology FIR learns inductively the rules of behavior of a system by analyzing input/output data alone.


Interested in reading the full paper? (18 pages, 112,177 bytes, pdf)


Homepage


Last modified: June 15, 2005 -- © François Cellier