Abstract
The problem of finding adequate population models in ecology is important for understanding essential aspects of their dynamic nature. Since analyzing and accurately predicting the intelligent adaptation of multiple species is difficult due to their complex interactions, the study of population dynamics still remains a challenging task in computational biology. In this paper, we use a modern deep reinforcement learning (RL) approach to explore a new avenue for understanding predator-prey ecosystems. Recently, reinforcement learning methods have achieved impressive results in areas, such as games and robotics. RL agents generally focus on building strategies for taking actions in an environment in order to maximize their expected returns. Here we frame the co-evolution of predators and preys in an ecosystem as allowing agents to learn and evolve toward better ones in a manner appropriate for multi-agent reinforcement learning. Recent significant advancements in reinforcement learning allow for new perspectives on these types of ecological issues. Our simulation results show that throughout the scenarios with RL agents, predators can achieve a reasonable level of sustainability, along with their preys.
Highlights
We model the learning problem for agents in the predator-prey ecosystem as a discounted reward reinforcement learning problem [24] with states s ∈ S, actions a ∈ A, discount rate γ ∈ (0, 1), and time steps t ∈ {0, 1, · · · }, in which learning agents interact with a Markov decision process (MDP) environment [48], which is defined by the tuple (S, A, T, r, γ)
We use a modern deep reinforcement learning (RL) approach to explore a new avenue for understanding key population dynamics of predator-prey ecosystems
Reinforcement learning methods have achieved impressive results, and reinforcement learning agents generally focus on building strategies that lead to agents taking actions in an environment in order to maximize expected reward
Summary
Many researchers have studied population models with evolutionary dispersal perspectives, such as dispersal depending on other species [1,2,3,4] and starvation-driven diffusion depending on resources [5,6,7,8,9,10]. Researchers have developed dispersal theory based on the surrounding environment as an influential element [12], and the environment affecting a particular species includes elements, such as other interacting species. Because various species usually migrate to a region to find a more favorable habitat that provides sufficient food and/or better conditions for survival, an understanding of dispersal strategy is critically important to the study of species evolution. For a general explanation of discrete and continuous models on dispersal evolution, we refer the reader to References [12,13,14,15,16]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
