🕤 Wednesday, 17th July 2024
Timings: Session 1 (09:30 to 11:00); Session 2 (11:30 to 13:00)
Speakers:
Manisha Padala
Mechanism Design and Deep Learning
Thirumalanathan D
Designing Optimal Auctions through Neural Networks
Mayank Ratan Bhardwaj
Fair Mechanism Design for Agriculture through Neural Networks
Shantanu Biswas
Building Fairness into AI: The Role of Mechanism Design
Abstract:
Mechanism Design, a game theoretic tool to elicit private information from strategic entities and aggregate it to achieve social good, has found many interesting applications in varied engineering fields. In computer science, applications such as ad auctions, crowdsourcing, crowdfunding, and Q&A forums are among the ones that are used.
In this tutorial, we will explain what mechanism design is. Designing an appropriate mechanism analytically is challenging with the increased complexity of modern AI applications. With deep neural networks’ success in learning many complex tasks in AI, researchers explored the possibility of a neural network learning mechanism for us. In this tutorial, we first learn mechanism design theory with illustrated examples, the basics of deep learning, and how it can be used in learning auctions. Next, we will learn about recent advancements in this domain for specific applications such as agriculture.
Timings: Session 1 (14:30 to 15:30); Session 2 (16:00 to 17:30)
Speakers:
Gogulapati Sreedurga
Introduction to Cooperative Game Theory
Swapnil Dhamal
Cooperative Game Theory and Influence Maximization
Ramasuri Narayanam
Cooperative Game Theory and Explainable AI
Abstract:
Cooperative game theory explores the formation of coalitions and the dynamics of cooperation among multiple agents. It employs mathematical methods to analyze scenarios where two or more agents collaborate to achieve specific goals. Recently, this paradigm has found application in diverse fields, such as Influence Maximization, Explainable AI (XAI), and Federated Learning (FL).
This tutorial aims to offer a comprehensive understanding of cooperative game theory, highlighting significant results in the field. Additionally, we will explore the role of cooperative game theory tools such as Shapley Value in cutting-edge AI applications and their competitive edge compared to alternative approaches.