| Project Detail |
"The field of Algorithmic Game Theory (AGT) has come of age. It has a rich theory, motivated by real-world scenarios, and has found increasingly widespread applications in electronic markets and online auctions. However, the state of the art of the AGT theory still lags far behind full applicability in many Internet applications, largely due to excessive model simplification, unsuitable analysis methods, and disregard of crucial real-world considerations. As increasingly more social and commerce applications move to the Internet, {\bf the robustness of AGT theory becomes a crucial issue}. The additional research needed to render AGT theory more suitable for the realm of practice in a more substantial way is far from negligible. However, the potential uses of this fundamental area are so significant that research in this direction is imperative. We foresee that the approaches pursued in this proposed research will play a central role in reaching this highly momentous goal.
To achieve this goal of advancing AGT theory, we will identify problematic assumptions and analysis methods employed in the traditional AGT paradigm, propose novel approaches for making these assumptions as flexible and nuanced as possible, and develop new analysis and design techniques to address more realistic scenarios. The proposed project will open up new horizons in the field of AGT, and will broaden its theoretical scope and real-world validity.
More than a decade ago, Papadimitriou wrote: ""A fascinating fusion of ideas from both fields --- game theory and algorithms --- came into being and was used productively in the effort to illuminate the mysteries of the Internet. It has come to be called algorithmic game theory."" My vision is that an impressive and cogent fusion of ideas will emerge from this project, that will come to be called ""robust algorithmic game theory"" (RAGT). |
