Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userDecentralized decision making by an ant colony: drift-diffusion model of individual choice, quorum and collective decision
82
0
0.0
(
0
)
Added by
Debashish Chowdhury
Publication date
2021
fields
Physics
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Ants are social insects. When the existing nest of an ant colony becomes uninhabitable, the hunt for a new suitable location for migration of the colony begins. Normally, multiple sites may be available as the potential new nest site. Distinct sites may be chosen by different scout ants based on their own assessments. Since the individual assessment is error prone, many ants may choose inferior site(s). But, the collective decision that emerges from the sequential and decentralized decision making process is often far better. We develop a model for this multi-stage decision making process. A stochastic drift-diffusion model (DDM) captures the sequential information accumulation by individual scout ants for arriving at their respective individual choices. The subsequent tandem runs of the scouts, whereby they recruit their active nestmates, is modelled in terms of suitable adaptations of the totally asymmetric simple exclusion processes (TASEP). By a systematic analysis of the model we explore the conditions that determine the speed of the emergence of the collective decision and the quality of that decision. More specifically, we demonstrate that collective decision of the colony is much less error-prone that the individual decisions of the scout ants. We also compare our theoretical predictions with experimental data.
rate research
Read More
An essential task of groups is to provide efficient solutions for the complex problems they face. Indeed, considerable efforts have been devoted to the question of collective decision-making related to problems involving a single dominant feature. Here we introduce a quantitative formalism for finding the optimal distribution of the group members competences in the more typical case when the underlying problem is complex, i.e., multidimensional. Thus, we consider teams that are aiming at obtaining the best possible answer to a problem having a number of independent sub-problems. Our approach is based on a generic scheme for the process of evaluating the proposed solutions (i.e., negotiation). We demonstrate that the best performing groups have at least one specialist for each sub-problem -- but a far less intuitive result is that finding the optimal solution by the interacting group members requires that the specialists also have some insight into the sub-problems beyond their unique field(s). We present empirical results obtained by using a large-scale database of citations being in good agreement with the above theory. The framework we have developed can easily be adapted to a variety of realistic situations since taking into account the weights of the sub-problems, the opinions or the relations of the group is straightforward. Consequently, our method can be used in several contexts, especially when the optimal composition of a group of decision-makers is designed.
The spreading dynamics of an epidemic and the collective behavioral pattern of the population over which it spreads are deeply intertwined and the latter can critically shape the outcome of the former. Motivated by this, we design a parsimonious game-theoretic behavioral--epidemic model, in which an interplay of realistic factors shapes the co-evolution of individual decision-making and epidemics on a network. Although such a co-evolution is deeply intertwined in the real-world, existing models schematize population behavior as instantaneously reactive, thus being unable to capture human behavior in the long term. Our model offers a unified framework to model and predict complex emergent phenomena, including successful collective responses, periodic oscillations, and resurgent epidemic outbreaks. The framework also allows to assess the effectiveness of different policy interventions on ensuring a collective response that successfully eradicates the outbreak. Two case studies, inspired by real-world diseases, are presented to illustrate the potentialities of the proposed model.
This paper seeks to establish a framework for directing a society of simple, specialized, self-interested agents to solve what traditionally are posed as monolithic single-agent sequential decision problems. What makes it challenging to use a decentralized approach to collectively optimize a central objective is the difficulty in characterizing the equilibrium strategy profile of non-cooperative games. To overcome this challenge, we design a mechanism for defining the learning environment of each agent for which we know that the optimal solution for the global objective coincides with a Nash equilibrium strategy profile of the agents optimizing their own local objectives. The society functions as an economy of agents that learn the credit assignment process itself by buying and selling to each other the right to operate on the environment state. We derive a class of decentralized reinforcement learning algorithms that are broadly applicable not only to standard reinforcement learning but also for selecting options in semi-MDPs and dynamically composing computation graphs. Lastly, we demonstrate the potential advantages of a societys inherent modular structure for more efficient transfer learning.
Algorithms engineered to leverage rich behavioral and biometric data to predict individual attributes and actions continue to permeate public and private life. A fundamental risk may emerge from misconceptions about the sensitivity of such data, as well as the agency of individuals to protect their privacy when fine-grained (and possibly involuntary) behavior is tracked. In this work, we examine how individuals adjust their behavior when incentivized to avoid the algorithmic prediction of their intent. We present results from a virtual reality task in which gaze, movement, and other physiological signals are tracked. Participants are asked to decide which card to select without an algorithmic adversary anticipating their choice. We find that while participants use a variety of strategies, data collected remains highly predictive of choice (80% accuracy). Additionally, a significant portion of participants became more predictable despite efforts to obfuscate, possibly indicating mistaken priors about the dynamics of algorithmic prediction.
Modern scientific research has become largely a cooperative activity in the Internet age. We build a simulation model to understand the population-level creativity based on the heuristic ant colony algorithm. Each researcher has two heuristic parameters characterizing the goodness of his own judgments and his trust on literature. In a population with all kinds of researchers, we find that as the problem scale increases, the contributor distribution significantly shifts from the independent regime of relying on ones own judgments to the cooperative regime of more closely following the literature. The distribution also changes with the stage of the research problem and the computing power available. Our work provides some preliminary understanding and guidance for the dynamical process of cooperative scientific research in various disciplines.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
