Title: Averaging processes on graphs with random initial opinions
Abstract: In this talk, we will consider two processes of evolving opinions on graphs: the edge-averaging process and the DeGroot dynamics. We will study the typical time it takes for the opinions to reach approximate consensus. In particular, we will show that when the initial opinions of the vertices are random, the convergence is much faster compared to deterministic (worst case) initial opinions.
Based on joint works with Yuval Peres and Ron Peretz.