Learned emergence in selfish collective motion

Abstract

To understand the collective motion of many individuals, we often rely on agent-based models with rules that may be computationally complex and involved. For biologically inspired systems in particular, this raises questions about whether the imposed rules are necessarily an accurate reflection of what is being followed. The basic premise of updating one’s state according to some underlying motivation is well suited to the realm of reservoir computing; however, entire swarms of individuals are yet to be tasked with learning movement in this framework. This work focuses on the specific case of many selfish individuals simultaneously optimizing their domains in a manner conducive to reducing their personal risk of predation. Using an echo state network and data generated from the agent-based model, we show that, with an appropriate representation of input and output states, this selfish movement can be learned. This suggests that a more sophisticated neural network, such as a brain, could also learn this behavior and provides an avenue to further the search for realistic movement rules in systems of autonomous individuals.

Reservoir computing with swarms

“…The ABM and RC modelling paradigms have similar and complementary approaches to processing information. Both were integrated for the first time in 12 , where trajectories generated from the ABM provided training data to the RC and the trained output from the RC was then used to update the agent's state in a separate ABM, ultimately demonstrating that the proposed agent-based movement rules were learnable. Here, we take advantage of the high level of transparency in ABM, which allows for a detailed and quantitative study of the relationship between the dynamics of the swarm and its function as a reservoir.…”

Applicability of the Future State Maximization Paradigm to Agent-Based Modeling: A Case Study on the Emergence of Socially Sub-Optimal Mobility Behavior

“…Recently, the idea of incorporating more refined algorithms has gained traction [20][21][22][23]. The past decade has brought about vast advances in the field of artificial intelligence (AI): ABM is often considered a part of this field [24], and it is, therefore, reasonable to consider other advances within this field as possible remedies for a lack of simulated intelligence in the agents.…”

“…Machine learning algorithms have gained popularity in many research areas: as such, they have also been incorporated in ABM, to achieve artificial intelligence in agents [20,22,23]. At the same time, the concept of bounded rationality is still rarely used in AI studies [16].…”

Combining machine learning and data assimilation to forecast dynamical systems from noisy partial observations

“…The training of random feature map networks only requires linear least-square regression and it was proven rigorously that random feature maps enjoy the so called universal approximation property which states that they can approximate any continuous function arbitrarily close (Park and Sandberg, 1991;Cybenko, 1989;Barron, 1993). The framework of random feature maps was extended to include internal dynamics in so called echo-state networks and reservoir computers with remarkable success in forecasting dynamical systems (Maass et al, 2002;Jaeger, 2002;Jaeger and Haas, 2004;Pathak et al, 2018a;Algar et al, 2019;Nadiga, 2020;Bollt, 2021;Gauthier et al, 2021;Platt et al, 2021).…”