Home About us Contact
|
Home About us Contact | ||
|
|
||
Long-term Predictions (long-term + prediction)
Selected AbstractsFood web complexity and chaotic population dynamicsECOLOGY LETTERS, Issue 3 2002Gregor F. Fussmann Abstract In mathematical models, very simple communities consisting of three or more species frequently display chaotic dynamics which implies that long-term predictions of the population trajectories in time are impossible. Communities in the wild tend to be more complex, but evidence for chaotic dynamics from such communities is scarce. We used supercomputing power to test the hypothesis that chaotic dynamics become less frequent in model ecosystems when their complexity increases. We determined the dynamical stability of a universe of mathematical, nonlinear food web models with varying degrees of organizational complexity. We found that the frequency of unpredictable, chaotic dynamics increases with the number of trophic levels in a food web but decreases with the degree of complexity. Our results suggest that natural food webs possess architectural properties that may intrinsically lower the likelihood of chaotic community dynamics. [source] Beating the random walk in Central and Eastern EuropeJOURNAL OF FORECASTING, Issue 3 2005Jesús Crespo Cuaresma Abstract We compare the accuracy of vector autoregressive (VAR), restricted vector autoregressive (RVAR), Bayesian vector autoregressive (BVAR), vector error correction (VEC) and Bayesian error correction (BVEC) models in forecasting the exchange rates of five Central and Eastern European currencies (Czech Koruna, Hungarian Forint, Slovak Koruna, Slovenian Tolar and Polish Zloty) against the US Dollar and the Euro. Although these models tend to outperform the random walk model for long-term predictions (6 months ahead and beyond), even the best models in terms of average prediction error fail to reject the test of equality of forecasting accuracy against the random walk model in short-term predictions. Copyright © 2005 John Wiley & Sons, Ltd. [source] Modeling volatile isoprenoid emissions , a story with split endsPLANT BIOLOGY, Issue 1 2008R. Grote Abstract Accurate prediction of plant-generated volatile isoprenoid fluxes is necessary for reliable estimation of atmospheric ozone and aerosol formation potentials. In recent years, significant progress has been made in understanding the environmental and physiological controls on isoprenoid emission and in scaling these emissions to canopy and landscape levels. We summarize recent developments and compare different approaches for simulating volatile isoprenoid emission and scaling up to whole forest canopies with complex architecture. We show that the current developments in modeling volatile isoprenoid emissions are "split-ended" with simultaneous but separated efforts in fine-tuning the empirical emission algorithms and in constructing process-based models. In modeling volatile isoprenoid emissions, simplified leaf-level emission algorithms (Guenther algorithms) are highly successful, particularly after scaling these models up to whole regions, where the influences of different ecosystem types, ontogenetic stages, and variations in environmental conditions on emission rates and dynamics partly cancel out. However, recent experimental evidence indicates important environmental effects yet unconsidered and emphasize, the importance of a highly dynamic plant acclimation in space and time. This suggests that current parameterizations are unlikely to hold in a globally changing and dynamic environment. Therefore, long-term predictions using empirical algorithms are not necessarily reliable. We show that process-based models have large potential to capture the influence of changing environmental conditions, in particular if the leaf models are linked with physiologically based whole-plant models. This combination is also promising in considering the possible feedback impacts of emissions on plant physiological status such as mitigation of thermal and oxidative stresses by volatile isoprenoids. It might be further worth while to incorporate main features of these approaches in regional empirically-based emission estimations thereby merging the "split ends". [source] Bootstrapping Cognition from Behavior,A Computerized Thought ExperimentCOGNITIVE SCIENCE - A MULTIDISCIPLINARY JOURNAL, Issue 3 2008Ralf Möller Abstract We show that simple perceptual competences can emerge from an internal simulation of action effects and are thus grounded in behavior. A simulated agent learns to distinguish between dead ends and corridors without the necessity to represent these concepts in the sensory domain. Initially, the agent is only endowed with a simple value system and the means to extract low-level features from an image. In the interaction with the environment, it acquires a visuo-tactile forward model that allows the agent to predict how the visual input is changing under its movements, and whether movements will lead to a collision. From short-term predictions based on the forward model, the agent learns an inverse model. The inverse model in turn produces suggestions about which actions should be simulated in long-term predictions, and long-term predictions eventually give rise to the perceptual ability. [source] | ||||||||||