Chapter 10 . Aigal Species Succession in Rivers
209
science. For example, although the results of this study can stand alone by their
good performance, they also can serve as information for developing Cascade
Artificial Neural Networks (CANN), linking together the various study sites (e.g.
between upper and lower river segments). Neuro-Genetic Learning (NGL), which
evolves either neural network's architecture or its weights, is another example for
using informatics in the prediction of phytoplankton dynamics in a time series
(Jeong et al. 2001b). With its sophisticated methods, ecological informatics is
highly suitable for searching out and predicting ecosystem dynamics.
Limnological
Hierarchy
Ba in a eragc
(Chlorophyll)
Ecological
Informatics
Hybrid Modelling
-
Modularization,
Parallel Proce ing,
ete.
Fuzzy Logie, ellular Automata
Artificial cural etworks
· volut ionary omputation
Adaptive Agent
Fig. 10.5. Contribution of Ecological Informatics to Hybrid Modelling of
Limnological Phenomena at Different Levels of Organisation