Collaborative Decision Making in Complex Adaptive Systems

Speaker:  Ramasuri Narayanam – Bangalore, India
Topic(s):  Computational Theory, Algorithms and Mathematics

Abstract

Collaborative decision making in complex adaptive systems refers to an analytic viewpoint that looks beyond an individual’s cognition to include strategic interactions of individuals with others (including non-human cognitive agents). Here the objective is to develop a far superior collective intelligence through computational modeling of strategic interactions among the agents. 

Collaborative decision making is an interdisciplinary research agenda spanning across Multi-agent systems, Game theory and Mechanism design, Theory of Mind models, Socio-Cognitive Architecture, Optimization, Complex adaptive networks, Reinforcement and Machine learning. The primary goal of this lecture is to develop the conceptual underpinnings of forming collaborative decisions among autonomous multi-agents in complex systems by designing novel game theoretic models. Initially this lecture provides a brief overview of various popular models for collective decision making settings. Then, in particular, this lecture deals with the process of presenting specific game theoretic models for the following scenario. Consider that there exists a set of autonomous agents or human experts. Each agent initially possesses a strategy, which potentially represents "a piece of information" such as a rank order on a collection of entities/individuals. The various strategies available to the agents come from a fixed discrete set and that the agents have different intrinsic preferences among these strategies. Consider that there exists a distance function on the strategy set that allows to compare any pair of strategies by measuring similarity/dissimilarity among strategies. Further, each agent's payoff is dependent not only on his/her chosen strategy, but also on the strategies chosen by its neighbor agents. These human experts or autonomous agents engage in repeated interactions and their opinions evolve as the agents update their strategies during the interactions. This process eventually leads to achieve either consensus among the agents or to minimize the disagreement in their opinions.

The game theoretic framework requires sensible organization of best-response dynamics in order to monitor how the agents' update their strategies leading to a solution that achieves either consensus among the agents or to minimize the disagreement in their opinions. This problem setting has several interesting business as well as real-life use-cases such as candidate hiring and price prediction of raw materials. 

About this Lecture

Number of Slides:  60
Duration:  60 minutes
Languages Available:  English
Last Updated: 

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.