• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 2
  • 1
  • Tagged with
  • 3
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Modelos de ecossistemas com interações não lineares. / Model ecosystems with nonlinear interspecies interactions.

Santos, Danielle Oliveira Costa 24 September 2004 (has links)
Neste trabalho investigamos as propriedade estatísticas de um modelo de coevolução de N espécies, sob a perspectiva da dinâmica de replicadores. As interações entre pares de espécies são dadas por variáveis aleatórias independentes, fixas no tempo. As interações são também simétricas, de modo que a dinâmica maximiza uma função de Lyapunov (o funcional adaptabilidade). Isto permite usar as técnicas da mecânica estatística de sistemas desordenados para determinar analiticamente as propriedades estatísticas dos estados estacionários, particularmente a diversidade de espécies (total de espécies coexistindo em um sistema ecológico). As auto-interações são iguais a um parâmetro de controle que mede a competição entre indivíduos de uma mesma espécie (competição intraespecífica). A cada espécie associamos um conjunto de p traços ou características, representados por variáveis binárias aleatórias distribuídas com igual probabilidade. As forças de interação são dadas por funções não lineares da regra de Hebb. Estas são funções moduladoras do número de elementos complementares entre os conjuntos de traços de um dado par de espécies. Estudamos analítica e numericamente o caso em que p é proporcional ao total de espécies na comunidade, via método de réplicas. A análise é possível devido ao resultado de Sompolinsky: funções não lineares da regra de Hebb são equivalentes, no limite de p extensivo, a regra de Hebb somada a um ruído gaussiano estático, cuja variância depende da forma da função moduladora. A competição intraespecífica, o total de traços, a presença de espécies altamente complementares e finalmente o peso dos termos de competição interespecífica (elementos não diagonais da matriz de acoplamentos) são as principais influências sobre o comportamento das grandezas termodinâmicas no equilíbrio, principalmente a diversidade. Os resultados analíticos concordam com a solução numérica da equação de replicadores, no regime em que as soluções de réplicas simétricas são estáveis. / We investigate the statistical properties of a coevolution model of N species using the replicator dynamics framework. The pairwise species interactions are given by independent quenched random variables. They are also symmetric, so that the dynamics maximizes a quadratic Lyapunov function (the fitness functional). This allows the use of tools of statistical mechanics of disordered systems to analyze the statistical properties of the equilibrium states, especially the ecosystem diversity (total number of coexisting species in an ecological system). The self-interactions are equal to a control parameter measuring the intraspecies competition. We associate to each species a set of p traits and represent them by independent random variables, equally distributed. The strength of the pairwise interactions is given by nonlinear functions of the Hebb rule. These are modulating functions of the number of complementary elements in the sets of traits of a given species pair. We study analytically and numerically the limit of extensive p, using the replica trick. The analytical approach is possible due to a result derived by Sompolinsky: in the limit of extensive p, nonlinear functions of the Hebb rule are equivalent to the Hebb rule plus a Gaussian static noise, whose variance is dependent on the form of the modulating function. The intraspecies competition, the total number of traits, the presence of highly complementary species pairs and the contribution of the nondiagonal elements of the interaction matrix are the main influences over the behavior of the equilibrium properties, principally the diversity. Our analytical results agree with the numerical solutions of the replicator equation in the regime of stable replica symmetric solutions.
2

Modelos de ecossistemas com interações não lineares. / Model ecosystems with nonlinear interspecies interactions.

Danielle Oliveira Costa Santos 24 September 2004 (has links)
Neste trabalho investigamos as propriedade estatísticas de um modelo de coevolução de N espécies, sob a perspectiva da dinâmica de replicadores. As interações entre pares de espécies são dadas por variáveis aleatórias independentes, fixas no tempo. As interações são também simétricas, de modo que a dinâmica maximiza uma função de Lyapunov (o funcional adaptabilidade). Isto permite usar as técnicas da mecânica estatística de sistemas desordenados para determinar analiticamente as propriedades estatísticas dos estados estacionários, particularmente a diversidade de espécies (total de espécies coexistindo em um sistema ecológico). As auto-interações são iguais a um parâmetro de controle que mede a competição entre indivíduos de uma mesma espécie (competição intraespecífica). A cada espécie associamos um conjunto de p traços ou características, representados por variáveis binárias aleatórias distribuídas com igual probabilidade. As forças de interação são dadas por funções não lineares da regra de Hebb. Estas são funções moduladoras do número de elementos complementares entre os conjuntos de traços de um dado par de espécies. Estudamos analítica e numericamente o caso em que p é proporcional ao total de espécies na comunidade, via método de réplicas. A análise é possível devido ao resultado de Sompolinsky: funções não lineares da regra de Hebb são equivalentes, no limite de p extensivo, a regra de Hebb somada a um ruído gaussiano estático, cuja variância depende da forma da função moduladora. A competição intraespecífica, o total de traços, a presença de espécies altamente complementares e finalmente o peso dos termos de competição interespecífica (elementos não diagonais da matriz de acoplamentos) são as principais influências sobre o comportamento das grandezas termodinâmicas no equilíbrio, principalmente a diversidade. Os resultados analíticos concordam com a solução numérica da equação de replicadores, no regime em que as soluções de réplicas simétricas são estáveis. / We investigate the statistical properties of a coevolution model of N species using the replicator dynamics framework. The pairwise species interactions are given by independent quenched random variables. They are also symmetric, so that the dynamics maximizes a quadratic Lyapunov function (the fitness functional). This allows the use of tools of statistical mechanics of disordered systems to analyze the statistical properties of the equilibrium states, especially the ecosystem diversity (total number of coexisting species in an ecological system). The self-interactions are equal to a control parameter measuring the intraspecies competition. We associate to each species a set of p traits and represent them by independent random variables, equally distributed. The strength of the pairwise interactions is given by nonlinear functions of the Hebb rule. These are modulating functions of the number of complementary elements in the sets of traits of a given species pair. We study analytically and numerically the limit of extensive p, using the replica trick. The analytical approach is possible due to a result derived by Sompolinsky: in the limit of extensive p, nonlinear functions of the Hebb rule are equivalent to the Hebb rule plus a Gaussian static noise, whose variance is dependent on the form of the modulating function. The intraspecies competition, the total number of traits, the presence of highly complementary species pairs and the contribution of the nondiagonal elements of the interaction matrix are the main influences over the behavior of the equilibrium properties, principally the diversity. Our analytical results agree with the numerical solutions of the replicator equation in the regime of stable replica symmetric solutions.
3

Micro-Foundations of Organizational Adaptation : A Field Study in the Evolution of Product Development Capabilities in a Design Firm

Salvato, Carlo January 2006 (has links)
The aim of this dissertation is to improve knowledge of how organizations adapt to their dynamic environments, by developing a detailed understanding of the configuration and evolution of organizational replicators. Among open questions in the literature on organizational adaptation, I have explored the following: How can the structure of organizational replicators and the nature of their components be realistically described? How do different organizational replicators interact with each other at different levels within and across organizational boundaries? How do replicators evolve? Why do firms need dynamic capabilities? I’ve addressed these questions by means of an embedded, longitudinal field study of Alessi, an Italian firm founded in 1921, active in the development and production of hundreds of design household products. Data analysis has been carried out in two steps. First, a longitudinal analysis of available primary and archival data has provided an in-depth understanding of the composite nature of organizational replicators, their mutual relationships, their evolution, their outcome stability. Second, a more structured investigation relying on Optimal Matching Analysis allowed to reliably develop an understanding of replicators complexity and of the mechanisms behind their evolution. There are four key findings. First, replicators are not simply behavioral entities—routines in the “narrow sense”. Reliable performance of a capability requires additional elements of physical, intellectual and social capital, which are essential components of replicators (or “Replication Base—RB”, as I suggest to label these more articulated organizational traits). Second, interactions among components of Replication Bases at different levels within and outside the organization suggest a more articulated perspective on how organizational knowledge develops. Components of Replication Bases are often located at different positions within the organization. Over time, knowledge of a particular organizational process takes different forms across the organizational hierarchy. What is local search at one level of analysis, gradually becomes sophisticated foresight at different, typically higher, levels. Third, over time Replication Bases evolve by means of a complex interplay between random mutations and intentional interventions, supported by articulated learning processes. Finally, development of higher-level replicators is not the ultimate answer to the challenge of adaptation. Rather, it allows managers to focus their intentional interventions to the higher-level problems posed by the dynamism of competitive environments. Part of this liberated managerial attention and resources are focused on the crucial, non-routine task of understanding how the organization’s idiosyncratic attributes affect its prospects in the specific competitive context. Taken together, these findings outline the microfoundations of a framework for interpreting organizational adaptation.

Page generated in 0.0714 seconds