I am the student of M.tech CSE, currently working on my Dissertation. I am using the concept of reinforcement learning in order to improve coordination among agents in my research work. My work will be based on two types of learners that will use the observation mechanism to learn the behavior of agents in 4 types of networks i.e random, scale-free, small world and ring. I am thinking to implement this strategy in NetLogo, for Simulation and implementation.I just want to confirm that, Is it possible to do reinforcement learning in NetLogo. As I have checked the models in the netlogo library but could not find any model based on reinforcement learning.Please help me if somebody is working on this strategy.

Ginni, this does sound like a use case for NetLogo. You will need to model
your learning, of course. Suppose learning is modeled by the probability of
succeeding at a certain task, so it's a number between 0 and 1. This number
would be entered into a turtles-own variable. Then, depending on what
agents do, that variable is updated. This would model your learning.

Visit the repository in my web page
http://www.diegoleal.info/repository.html. You will be able to find the
NetLogo implementation of a prominent sociological article that uses
stochastic learning and adaptive thresholds