Text only print
Full screen print
Tokyo Probability Seminar
Seminar information archive ～06/12｜Next seminar｜Future seminars 06/13～
| Date, time & place | Monday 16:00 - 17:30 126Room #126 (Graduate School of Math. Sci. Bldg.) |
|---|---|
| Organizer(s) | Makiko Sasada, Shuta Nakajima, Masato Hoshino |
2025/06/02
16:00-17:30 Room #126 (Graduate School of Math. Sci. Bldg.)
We are having teatime from 15:15 in the common room on the second floor. Please join us.
Wai-Kit Lam (National Taiwan University)
Disorder monomer-dimer model and maximum weight matching
We are having teatime from 15:15 in the common room on the second floor. Please join us.
Wai-Kit Lam (National Taiwan University)
Disorder monomer-dimer model and maximum weight matching
[ Abstract ]
Given a finite graph, one puts i.i.d. weights on the edges and i.i.d. weights on the vertices. For a (partial) matching on this graph, define the weight of the matching by adding all the weights of the edges in the matching together with the weights of the unmatched vertices. One would like to understand how the maximum weight behaves as the size of the graph becomes large. The talk will be divided into two parts. In the first part, we consider the "positive temperature" case (a.k.a. the disorder monomer-dimer model). We show that the model exhibits correlation decay, and from this one can prove a Gaussian central limit theorem for the associated free energy. In the second part, we will focus on the "zero temperature" case, the maximum weight matching. We show that if the edge weights are exponentially distributed, and if the vertex weights are absent, then there is also correlation decay for a certain class of graphs. This correlation decay allows us to define the maximum weight matching on some infinite graphs and also prove limit theorems for the maximum weight matching. Joint work with Arnab Sen (Minnesota).
Given a finite graph, one puts i.i.d. weights on the edges and i.i.d. weights on the vertices. For a (partial) matching on this graph, define the weight of the matching by adding all the weights of the edges in the matching together with the weights of the unmatched vertices. One would like to understand how the maximum weight behaves as the size of the graph becomes large. The talk will be divided into two parts. In the first part, we consider the "positive temperature" case (a.k.a. the disorder monomer-dimer model). We show that the model exhibits correlation decay, and from this one can prove a Gaussian central limit theorem for the associated free energy. In the second part, we will focus on the "zero temperature" case, the maximum weight matching. We show that if the edge weights are exponentially distributed, and if the vertex weights are absent, then there is also correlation decay for a certain class of graphs. This correlation decay allows us to define the maximum weight matching on some infinite graphs and also prove limit theorems for the maximum weight matching. Joint work with Arnab Sen (Minnesota).
