Date:  Thursday, June 13th, 2019

Time:  10: 30 am – 12:00 pm

Location: 109, Essex Hall

 

Multi-Agent Based Modeling helps in simulating the actions and interactions of the intelligent agents in natural systems. In our research work, we will focus on simulating the evolution of agent capabilities and to make an intelligent agent specialized in terms of task allocation by considering the positive social influence in a given dynamic environment. A socially complex system can be seen as a social network of interacting Consumer-Agents and Producer-Agents. The goal of an agent being a desired state that it intends to reach, a composite goal hierarchy – We implement the goal selection algorithm in a dynamic environment. Initially, Consumer-Agents have minimal capabilities and need to acquire the required capabilities to achieve their goals. Different Producer-Agents extend their support by providing a variety of services for a fixed operational cost. In our thesis, we propose a novel method which exploits the benefits of social networks in a cultural evolutionary component and provides a decision support system to aid the Consumer-Agents in learning required capabilities from the best available Producer-Agents with minimum operational cost and to evolve into specialized agents by using the resource allocation techniques on selected tasks. We aim at presenting an augmented model for the evolution of specialized agents in a given dynamic environment. Additionally, our work demonstrates a significant reduction in the learning time, process time and operational costs.

Thesis Committee:

MSc Thesis Proposal Announcement

 

5113 Lambton Tower, 401 Sunset. Ave., Windsor ON, N9B 3P4 (519)253-3000, Ext. 3716, csgradinfo@uwindsor.ca