A rede de marcação e recaptura: o caso de borboletas do gênero heliconius

(The Mark and Recapture Network: a Heliconius case study). The current pace of habitat destruction, especially in tropical landscapes, has increased the need for understanding minimum patch requirements and patch distance as tools for conserving species in forest remnants. Mark recapture and tagging studies have been instrumental in providing parameters for functional models. Because of their popularity, ease of manipulation and well known biology, butterflies have become model in studies of spatial structure. Yet, most studies on butterflies movement have focused on temperate species that live in open habitats, in which forest patches are barrier to movement. This study aimed to view and review data from mark-recapture as a network in two species of butterfly (Heliconius erato and Heliconius melpomene). A work of marking and recapture of the species was carried out in an Atlantic forest reserve located about 20km from the city of Natal (RN). Mark recapture studies were conducted in 3 weekly visits during January-February and July-August in 2007 and 2008. Captures were more common in two sections of the dirt road, with minimal collection in the forest trail. The spatial spread of captures was similar in the two species. Yet, distances between recaptures seem to be greater for Heliconius erato than for Heliconius melpomene. In addition, the erato network is more disconnected, suggesting that this specie has shorter traveling patches. Moving on to the network, both species have similar number of links (N) and unweighed vertices (L). However, melpomene has a weighed network 50% more connections than erato. These network metrics suggest that erato has more compartmentalized network and restricted movement than melpomene. Thus, erato has a larger number of disconnected components, nC, in the network, and a smaller network diameter. The frequency distribution of network connectivity for both species was better explained by a Power-law than by a random, Poissom distribution, showing that the Power-law provides a better fit than the Poisson for both species. Moreover, the Powerlaw erato is much better adjusted than in melpomene, which should be linked to the small movements that erato makes in the network

Análise e desenvolvimento de controladores preditivos multivariáveis baseados em multi-modelos bilineares

This work addresses issues related to analysis and development of multivariable predictive controllers based on bilinear multi-models. Linear Generalized Predictive Control (GPC) monovariable and multivariable is shown, and highlighted its properties, key features and applications in industry. Bilinear GPC, the basis for the development of this thesis, is presented by the time-step quasilinearization approach. Some results are presented using this controller in order to show its best performance when compared to linear GPC, since the bilinear models represent better the dynamics of certain processes. Time-step quasilinearization, due to the fact that it is an approximation, causes a prediction error, which limits the performance of this controller when prediction horizon increases. Due to its prediction error, Bilinear GPC with iterative compensation is shown in order to minimize this error, seeking a better performance than the classic Bilinear GPC. Results of iterative compensation algorithm are shown. The use of multi-model is discussed in this thesis, in order to correct the deficiency of controllers based on single model, when they are applied in cases with large operation ranges. Methods of measuring the distance between models, also called metrics, are the main contribution of this thesis. Several application results in simulated distillation columns, which are close enough to actual behaviour of them, are made, and the results have shown satisfactory

Contribuições aos Processos de Clustering com Base em Métricas não-Euclidianas

In this work we present a new clustering method that groups up points of a data set in classes. The method is based in a algorithm to link auxiliary clusters that are obtained using traditional vector quantization techniques. It is described some approaches during the development of the work that are based in measures of distances or dissimilarities (divergence) between the auxiliary clusters. This new method uses only two a priori information, the number of auxiliary clusters Na and a threshold distance dt that will be used to decide about the linkage or not of the auxiliary clusters. The number os classes could be automatically found by the method, that do it based in the chosen threshold distance dt, or it is given as additional information to help in the choice of the correct threshold. Some analysis are made and the results are compared with traditional clustering methods. In this work different dissimilarities metrics are analyzed and a new one is proposed based on the concept of negentropy. Besides grouping points of a set in classes, it is proposed a method to statistical modeling the classes aiming to obtain a expression to the probability of a point to belong to one of the classes. Experiments with several values of Na e dt are made in tests sets and the results are analyzed aiming to study the robustness of the method and to consider heuristics to the choice of the correct threshold. During this work it is explored the aspects of information theory applied to the calculation of the divergences. It will be explored specifically the different measures of information and divergence using the Rényi entropy. The results using the different metrics are compared and commented. The work also has appendix where are exposed real applications using the proposed method

Generalizações do conceito de distância, i-distâncias, distâncias intervalares e topologia

In this dissertation we present some generalizations for the concept of distance by using more general value spaces, such as: fuzzy metrics, probabilistic metrics and generalized metrics. We show how such generalizations may be useful due to the possibility that the distance between two objects could carry more information about the objects than in the case where the distance is represented just by a real number. Also in this thesis we propose another generalization of distance which encompasses the notion of interval metric and generates a topology in a natural way. Several properties of this generalization are investigated, and its links with other existing generalizations

Algoritmos de agrupamentos fuzzy intervalares e índice de validação para agrupamento de dados simbólicos do tipo intervalo

Symbolic Data Analysis (SDA) main aims to provide tools for reducing large databases to extract knowledge and provide techniques to describe the unit of such data in complex units, as such, interval or histogram. The objective of this work is to extend classical clustering methods for symbolic interval data based on interval-based distance. The main advantage of using an interval-based distance for interval-based data lies on the fact that it preserves the underlying imprecision on intervals which is usually lost when real-valued distances are applied. This work includes an approach allow existing indices to be adapted to interval context. The proposed methods with interval-based distances are compared with distances punctual existing literature through experiments with simulated data and real data interval

Contribuição ao estudo de redes complexas: modelo de afinidade com métrica

Currently the interest in large-scale systems with a high degree of complexity has been much discussed in the scientific community in various areas of knowledge. As an example, the Internet, protein interaction, collaboration of film actors, among others. To better understand the behavior of interconnected systems, several models in the area of complex networks have been proposed. Barabási and Albert proposed a model in which the connection between the constituents of the system could dynamically and which favors older sites, reproducing a characteristic behavior in some real systems: connectivity distribution of scale invariant. However, this model neglects two factors, among others, observed in real systems: homophily and metrics. Given the importance of these two terms in the global behavior of networks, we propose in this dissertation study a dynamic model of preferential binding to three essential factors that are responsible for competition for links: (i) connectivity (the more connected sites are privileged in the choice of links) (ii) homophily (similar connections between sites are more attractive), (iii) metric (the link is favored by the proximity of the sites). Within this proposal, we analyze the behavior of the distribution of connectivity and dynamic evolution of the network are affected by the metric by A parameter that controls the importance of distance in the preferential binding) and homophily by (characteristic intrinsic site). We realized that the increased importance as the distance in the preferred connection, the connections between sites and become local connectivity distribution is characterized by a typical range. In parallel, we adjust the curves of connectivity distribution, for different values of A, the equation P(k) = P0e

When wolves cry: long distance calling by wild maned wolves

The maned wolf (Chrysocyon brachyurus Illiger 1815) is the biggest canid in South America and it is considered a “near threatened” species by IUCN. Because of its nocturnal, territorial and solitary habits, there are still many understudied aspects of their behavior in natural environments, including acoustic communication. In its vocal repertoire, the wolf presents a longdistance call named “roar-bark” which, according to literature, functions for spacing maintenance between individuals and/or communication between members of the reproductive pair inside the territory. In this context, this study aimed: 1) to compare four methods for detecting maned wolf’s roar-barks in recordings made in a natural environment, in order to elect the most efficient one for our project; 2) to understand the night emission pattern of these vocalizations, verifying possible weather and moon phases influences in roarbark’s emission rates; and 3) to test Passive Acoustic Monitoring as a tool to identify the presence of maned wolves in a natural environment. The study area was the Serra da Canastra National Park (Minas Gerais, Brazil), where autonomous recorders were used for sound acquisition, recording all night (from 06pm to 06am) during five days in December/2013 and every day from April to July/2014. Roar-barks’ detection methods were tested and compared regarding time needed to analyze files, number of false positives and number of correctly identified calls. The mixed method (XBAT + manual) was the most efficient one, finding 100% of vocalizations in almost half of the time the manual method did, being chosen for our data analysis. By studying roarbarks’ temporal variation we verified that the wolves vocalize more in the early hours of the evening, suggesting an important social function for those calls at the beginning of its period of most intense activity. Average wind speed negatively influenced vocalization rate, which may indicate lower sound reception of recorders or a change in behavioral patterns of wolves in high speed wind conditions. A better understanding of seasonal variation of maned wolves’ vocal activity is required, but our study already shows that it is possible to detect behavioral patterns of wild animals only by sound, validating PAM as a tool in this species’ conservation.