This model was developed to test the usability of evolutionary computing and reinforcement learning by extending a well known agent-based model. Sugarscape (Epstein & Axtell, 1996) has been used to demonstrate migration, trade, wealth inequality, disease processes, sex, culture, and conflict. It is on conflict that this model is focused to demonstrate how machine learning methodologies could be applied.

The code is based on the Sugarscape 2 Constant Growback model, availble in the NetLogo models library. New code was added into the existing model while removing code that was not needed and modifying existing code to support the changes. Support for the original movement rule was retained while evolutionary computing, Q-Learning, and SARSA Learning were added.

The agent-based simulation is set to work on information that is either (a) functional, (b) pseudo-functional, (c) dysfunctional, or (d) irrelevant. The idea is that a judgment on whether information falls into one of the four categories is based on the agent and its network. In other words, it is the agents who interprets a particular information as being (a), (b), (c), or (d). It is a decision based on an exchange with co-workers. This makes the judgment a socially-grounded cognitive exercise. The uFUNK 1.0.2 Model is set on an organization where agent-employee work on agent-tasks.

This repository the multi-agent simulation software for the paper “Comparison of Competing Market Mechanisms with Reinforcement Learning in a CarPooling Scenario”. It’s a mutlithreaded Javaapplication.

Overview

Purpose

Modeling an economy with stable macro signals, that works as a benchmark for studying the effects of the agent activities, e.g. extortion, at the service of the elaboration of public policies..
…

We model the epistemic dynamics preceding political uprising. Before deciding whether to start protests, agents need to estimate the amount of discontent with the regime. This model simulates the dynamics of group knowledge about general discontent.

The aim of this model is to explore and understand the factors driving adoption of treatment strategies for ecological disturbances, considering payoff signals, learning strategies and social-ecological network structure

This model implements a classic scenario used in Reinforcement Learning problem, the “Cliff Walking Problem”. Consider the gridworld shown below (SUTTON; BARTO, 2018). This is a standard undiscounted, episodic task, with start and goal states, and the usual actions causing movement up, down, right, and left. Reward is -1 on all transitions except those into the region marked “The Cliff.” Stepping into this region incurs a reward of -100 and sends the agent instantly back to the start (SUTTON; BARTO, 2018).

The problem is solved in this model using the Q-Learning algorithm. The algorithm is implemented with the support of the NetLogo Q-Learning Extension

This is a re-implementation of a the NetLogo model Maze (ROOP, 2006).

This re-implementation makes use of the Q-Learning NetLogo Extension to implement the Q-Learning, which is done only with NetLogo native code in the original implementation.

Exploring how learning and social-ecological networks influence management choice set and their ability to increase the likelihood of species coexistence (i.e. biodiversity) on a fragmented landscape controlled by different managers.

This NetLogo model illustrates the cultural evolution of pro-environmental behaviour patterns. It illustrates how collective behaviour patterns evolve from interactions between agents and agents (in a social network) as well as agents and the affordances (action opportunities provided by the environment) within a niche. More specifically, the cultural evolution of behaviour patterns is understood in this model as a product of:

1. The landscape of affordances provided by the material environment,
2. Individual learning and habituation,
3. Social learning and network structure,
4. Personal states (such as habits and attitudes), and
   …