#95 Identity, Kinship, and the Evolution of Cooperation
Authors: Burton Voorhees, Dwight Read and Liane Gabora
Extensive cooperation among large groups of strangers, going well beyond behavior that provides biological or socio-economic reward, is unique to humans. We are called “Supercooperators,” the “cooperative species.” Most theoretical work on the evolution of cooperation treats it as a biological phenomenon, but empirically humans are more cooperative than current theoretical explanatory mechanisms would suggest. In our view, this lack of theoretical match to empirical findings arises because the complex interactions of the social, cultural, and psychological aspects of cooperation are not adequately taken into account. This paper proposes a new approach to the study of the evolution of cooperation in human populations. We propose that this evolution required (1) the evolution of the capacity for self-sustained, self-referential thought manifesting as an integrated worldview, a sense of identity, and a point of view, and (2) the co-evolution of kinship-based organizational systems within which stable, cognized social identities arise. Thus, individuals have a personal identity by virtue of being a sentient organism, and a social identity by virtue of being a social being. Within the field established by this polarity, group boundaries, rules of behavior, beliefs, social norms, and other practices are culturally integrated. While current culture-gene coevolution theories attempt to explain human cooperative behavior in terms of genetically evolved cooperative tendencies, we place more emphasis on top down cultural factors. In our view, humans are biologically disinclined to cooperate with other than close biological kin or through direct reciprocity, but have developed social identities carrying cultural expectations of cooperation with all those recognized as cultural kin. That is, humans cooperate with non-biological kin in large part because an individual acts in accord with a cultural concept of “who I am and what obligations that imposes on me.”
#96 Neutral Space Associated to Boolean Microbial Interaction Networks
Authors: Dante Travisany, Alejandro Maass and Eric Goles
The interactions between microorganisms in a determined community (microbiome) give rise to complex circuits known as microbial interaction networks (MIN), these interactions provides resilience capacity and stability to the whole community. One of the most studied microbiome is the Human Gut Microbiome, which remains globally stable over time in healthy conditions and absence of perturbation. Yet, recurrent alterations such as antibiotic treatment could produce shifts in the bacterial composition providing a void filled by opportunistic pathogens. One example is the Clostridium difficile infection (CDI). To study the dynamics of the Human gut microbiome under antibiotic perturbation and CDI, Steinway et al. uses previously reported time series datasets to infer a Boolean MIN (CDI-BMIN). This group recovers the known dynamics of Clindamycin antibiotic treatment and CDI. Nevertheless, the topological robustness of this model has not yet been studied. One approach is through the neutral space. In this work, we construct a reduced threshold Boolean network model (TBN) of the CDI-BMIN. Then, we explore the neutral space through evolutionary computation to generate candidate solutions that have at least the same fixed points of the CDI-BMIN. Using this TBN as input of the evolutionary algorithm, we found more than 5000 new TBN. We filter the results to keep only the solutions that share exactly the same fixed points of the original CDI-BMIN. Then, we construct the neutral graph. To enrich the neutral graph, we applied the same strategy as before, but selecting 5 previously generated solutions as a new input for the algorithm. New TBNs were found and added to the neutral graph. The results show several differences between the connected component solutions of the neutral graph and the other solutions; differences in the maximum and minimum hamming distance. Also, the connected component of the neutral graph is a scale free network.
#99 Blending Complexity Research
Authors: Kunbei Zhang and Aernout Schmidt
Disruptive complex adaptive systems (CASs, e.g. the profiling/social-media complex) put more and more pressures on individual agents' capacities to act autonomously and responsible. As legal theorists we have some questions here. The foundation of our trade is a conception of individual freedom that supports liability for deliberate behaviors. Thus: pressures on individual agents' capacities to act autonomously disrupt the basic paradigm of our discipline. This is worth investigating. Yet, investigating the potential of CASs to disrupt the law as a cultural survival vehicle requires “blended research,” by a whole gamut of disciplines, including legal theory.
Blending different disciplinary efforts is notoriously difficult, as each discipline tends to develop its own technical language for its own specific specializations. `Tort' is a different concept for legal and for non-legal specialists, as `rational choice' is for economic and non-economic specialists, as `culture' is for anthropologists and non-anthropologists and as `inflation' is for cosmologists and non-cosmologists. These differences are sticky, as they tend to nurse specialist identity. In order to support diverse specialists to work together on complex problems in cross-disciplinary teams we need a lingua franca that allows for discussing and deciding on the merits of the models involved while leaving relevant parts of the domain of discourse to specialists. How can we do that?
That is what this contribution is about. We distill an instance of the lingua franca required from the requirements-engineering literature that has emerged in the computing sciences around 2000. We introduce simple techniques to support cross-disciplinary discussion, using Acemoglu and Robinson (2001)'s formal econometric model of democratic/non-democratic regime change as a vehicle. And we show how the approach helps improve access to comprehension of complex situations by blending cross-disciplinary contributions.
#113 Simultaneous estimation and selection of species diversification model.
Authors: Francisco Richter, Ernst Wit and Rampal Etienne
Biodiversity, the term used to describe the wide variety of species on Earth, is declining at enormous rates due to human-induced environmental changes. This compromises ecosystem stability and productivity, which negatively impacts the ecosystem services that support life on Earth.
The mechanisms that control the diversification of species are poorly understood. In the last decade sophisticated diversification models have been developed, but these models ignore ecological interactions. While current models have examined factors such as competition, predator-prey interactions, parasitism, and mutualism, they have been developed on a case-by-case basis, and no general method to study the combined effect of these factors exists.
Such a general method has remained elusive for several reasons. Firstly, evolutionary processes have extremely complex dynamics. Secondly, decay and fossilization degrade crucial evidence useful for phylogenetic analyses that could infer underlying mechanisms. Thirdly, diversification processes have many potential explanatory variables, which increases the dimensionality of the models enormously.
To overcome these issues, we propose a general speciation model with potentially many covariates. This complex stochastic differential equation model can be written equivalently as a combination of two generalized linear models. The fact that we typically only have data on currently existing species observe the current species can be described as a missing data problem. For sparse inference of the speciation parameters we integrate a sequential path estimator in the M-step of the EM algorithm. We make use of the underlying differential geometry of the speciation models using a sparse, computationally feasible and consistent model selection procedure (dgLARS). The dgLARS algorithm has been successfully applied in the context of sparse generalized linear models, but their application to diversification models is a novel methodological contribution.
We show that our model has the potential to select the ecological factors that are key to driving species diversification.
#114 How transfer flights shape the structure of the airline network
Authors: Tomasz Ryczkowski, Agata Fronczak and Piotr Fronczak
We analyse data coming from the International Passenger Flight Traffic in 1990-2011 and show, how the transfer flights affect the overall structure of the international airline connections. Different quantitative and numerical approaches underlying the field of complex networks (including the idea of exponential random graph models and agent-based modelling) are used to tackle the relation between the direct and transfer flights. The possible applications of our approach to problems such as proposing the location of new flight connections and new airports is also discussed.
Authors: Burton Voorhees, Dwight Read and Liane Gabora
Extensive cooperation among large groups of strangers, going well beyond behavior that provides biological or socio-economic reward, is unique to humans. We are called “Supercooperators,” the “cooperative species.” Most theoretical work on the evolution of cooperation treats it as a biological phenomenon, but empirically humans are more cooperative than current theoretical explanatory mechanisms would suggest. In our view, this lack of theoretical match to empirical findings arises because the complex interactions of the social, cultural, and psychological aspects of cooperation are not adequately taken into account. This paper proposes a new approach to the study of the evolution of cooperation in human populations. We propose that this evolution required (1) the evolution of the capacity for self-sustained, self-referential thought manifesting as an integrated worldview, a sense of identity, and a point of view, and (2) the co-evolution of kinship-based organizational systems within which stable, cognized social identities arise. Thus, individuals have a personal identity by virtue of being a sentient organism, and a social identity by virtue of being a social being. Within the field established by this polarity, group boundaries, rules of behavior, beliefs, social norms, and other practices are culturally integrated. While current culture-gene coevolution theories attempt to explain human cooperative behavior in terms of genetically evolved cooperative tendencies, we place more emphasis on top down cultural factors. In our view, humans are biologically disinclined to cooperate with other than close biological kin or through direct reciprocity, but have developed social identities carrying cultural expectations of cooperation with all those recognized as cultural kin. That is, humans cooperate with non-biological kin in large part because an individual acts in accord with a cultural concept of “who I am and what obligations that imposes on me.”
#96 Neutral Space Associated to Boolean Microbial Interaction Networks
Authors: Dante Travisany, Alejandro Maass and Eric Goles
The interactions between microorganisms in a determined community (microbiome) give rise to complex circuits known as microbial interaction networks (MIN), these interactions provides resilience capacity and stability to the whole community. One of the most studied microbiome is the Human Gut Microbiome, which remains globally stable over time in healthy conditions and absence of perturbation. Yet, recurrent alterations such as antibiotic treatment could produce shifts in the bacterial composition providing a void filled by opportunistic pathogens. One example is the Clostridium difficile infection (CDI). To study the dynamics of the Human gut microbiome under antibiotic perturbation and CDI, Steinway et al. uses previously reported time series datasets to infer a Boolean MIN (CDI-BMIN). This group recovers the known dynamics of Clindamycin antibiotic treatment and CDI. Nevertheless, the topological robustness of this model has not yet been studied. One approach is through the neutral space. In this work, we construct a reduced threshold Boolean network model (TBN) of the CDI-BMIN. Then, we explore the neutral space through evolutionary computation to generate candidate solutions that have at least the same fixed points of the CDI-BMIN. Using this TBN as input of the evolutionary algorithm, we found more than 5000 new TBN. We filter the results to keep only the solutions that share exactly the same fixed points of the original CDI-BMIN. Then, we construct the neutral graph. To enrich the neutral graph, we applied the same strategy as before, but selecting 5 previously generated solutions as a new input for the algorithm. New TBNs were found and added to the neutral graph. The results show several differences between the connected component solutions of the neutral graph and the other solutions; differences in the maximum and minimum hamming distance. Also, the connected component of the neutral graph is a scale free network.
#99 Blending Complexity Research
Authors: Kunbei Zhang and Aernout Schmidt
Disruptive complex adaptive systems (CASs, e.g. the profiling/social-media complex) put more and more pressures on individual agents' capacities to act autonomously and responsible. As legal theorists we have some questions here. The foundation of our trade is a conception of individual freedom that supports liability for deliberate behaviors. Thus: pressures on individual agents' capacities to act autonomously disrupt the basic paradigm of our discipline. This is worth investigating. Yet, investigating the potential of CASs to disrupt the law as a cultural survival vehicle requires “blended research,” by a whole gamut of disciplines, including legal theory.
Blending different disciplinary efforts is notoriously difficult, as each discipline tends to develop its own technical language for its own specific specializations. `Tort' is a different concept for legal and for non-legal specialists, as `rational choice' is for economic and non-economic specialists, as `culture' is for anthropologists and non-anthropologists and as `inflation' is for cosmologists and non-cosmologists. These differences are sticky, as they tend to nurse specialist identity. In order to support diverse specialists to work together on complex problems in cross-disciplinary teams we need a lingua franca that allows for discussing and deciding on the merits of the models involved while leaving relevant parts of the domain of discourse to specialists. How can we do that?
That is what this contribution is about. We distill an instance of the lingua franca required from the requirements-engineering literature that has emerged in the computing sciences around 2000. We introduce simple techniques to support cross-disciplinary discussion, using Acemoglu and Robinson (2001)'s formal econometric model of democratic/non-democratic regime change as a vehicle. And we show how the approach helps improve access to comprehension of complex situations by blending cross-disciplinary contributions.
#113 Simultaneous estimation and selection of species diversification model.
Authors: Francisco Richter, Ernst Wit and Rampal Etienne
Biodiversity, the term used to describe the wide variety of species on Earth, is declining at enormous rates due to human-induced environmental changes. This compromises ecosystem stability and productivity, which negatively impacts the ecosystem services that support life on Earth.
The mechanisms that control the diversification of species are poorly understood. In the last decade sophisticated diversification models have been developed, but these models ignore ecological interactions. While current models have examined factors such as competition, predator-prey interactions, parasitism, and mutualism, they have been developed on a case-by-case basis, and no general method to study the combined effect of these factors exists.
Such a general method has remained elusive for several reasons. Firstly, evolutionary processes have extremely complex dynamics. Secondly, decay and fossilization degrade crucial evidence useful for phylogenetic analyses that could infer underlying mechanisms. Thirdly, diversification processes have many potential explanatory variables, which increases the dimensionality of the models enormously.
To overcome these issues, we propose a general speciation model with potentially many covariates. This complex stochastic differential equation model can be written equivalently as a combination of two generalized linear models. The fact that we typically only have data on currently existing species observe the current species can be described as a missing data problem. For sparse inference of the speciation parameters we integrate a sequential path estimator in the M-step of the EM algorithm. We make use of the underlying differential geometry of the speciation models using a sparse, computationally feasible and consistent model selection procedure (dgLARS). The dgLARS algorithm has been successfully applied in the context of sparse generalized linear models, but their application to diversification models is a novel methodological contribution.
We show that our model has the potential to select the ecological factors that are key to driving species diversification.
#114 How transfer flights shape the structure of the airline network
Authors: Tomasz Ryczkowski, Agata Fronczak and Piotr Fronczak
We analyse data coming from the International Passenger Flight Traffic in 1990-2011 and show, how the transfer flights affect the overall structure of the international airline connections. Different quantitative and numerical approaches underlying the field of complex networks (including the idea of exponential random graph models and agent-based modelling) are used to tackle the relation between the direct and transfer flights. The possible applications of our approach to problems such as proposing the location of new flight connections and new airports is also discussed.
Home |
Calls |
Programme |
Registration |
Contact |
Copyright CCS © 2016