Design of neuro-fuzzy models by evolutionary and gradient-based algorithms

All systems found in nature exhibit, with different degrees, a nonlinear behavior. To emulate this behavior, classical systems identification techniques use, typically, linear models, for mathematical simplicity. Models inspired by biological principles (artificial neural networks) and linguistically motivated (fuzzy systems), due to their universal approximation property, are becoming alternatives to classical mathematical models. In systems identification, the design of this type of models is an iterative process, requiring, among other steps, the need to identify the model structure, as well as the estimation of the model parameters. This thesis addresses the applicability of gradient-basis algorithms for the parameter estimation phase, and the use of evolutionary algorithms for model structure selection, for the design of neuro-fuzzy systems, i.e., models that offer the transparency property found in fuzzy systems, but use, for their design, algorithms introduced in the context of neural networks. A new methodology, based on the minimization of the integral of the error, and exploiting the parameter separability property typically found in neuro-fuzzy systems, is proposed for parameter estimation. A recent evolutionary technique (bacterial algorithms), based on the natural phenomenon of microbial evolution, is combined with genetic programming, and the resulting algorithm, bacterial programming, advocated for structure determination. Different versions of this evolutionary technique are combined with gradient-based algorithms, solving problems found in fuzzy and neuro-fuzzy design, namely incorporation of a-priori knowledge, gradient algorithms initialization and model complexity reduction.

Herbivores strongly influence algal recruitment in both coral- and algal-dominated coral reef habitats

Coral reefs can exist as coral- and macroalgae-dominated habitats often separated by only a few hundred metres. While herbivorous fish are known to depress the abundance of algae and help maintain the function of coral-dominated habitats, less is known about their influence in algae-dominated habitats. Here, we quantified herbivorous fish and benthic algal communities over a 6 mo period in coral-dominated (back-reef) and algal-dominated (lagoon) habitats in a relatively undisturbed fringing coral reef (Ningaloo, Western Australia). Simulta - neously, we tested the effects of herbivorous fish on algal recruitment in both habitats using recruitment tiles and fish exclusion cages. The composition of established algal communities differed consistently between habitats, with the back-reef hosting a more diverse community than the Sargassum-dominated lagoon. However, total algal biomass and cover only differed between habitats in autumn, coinciding with maximum Sargassum biomass. The back-reef hosted high coral cover and a diverse herbivorous fish community, with herbivore biomass an order of magnitude greater than the lagoon. Despite these differences in herbivore composition, exclusion of large herbivores had a similar positive effect to foliose macroalgae recruitment on experimental tiles in both back-reef and lagoon habitats. Additionally, territorial damselfish found in the backreef increased turf algae cover and decreased crustose coralline algae cover on recruitment tiles. Collectively, our results show that disparate herbivorous fish communities in coral- and algaedominated habitats are similarly able to limit the recruitment of foliose macroalgae, but suggest that when herbivorous fish biomass and diversity are relatively low, macroalgal communities are able to escape herbivore control through increased growth.