LuanVan.vnLuận Văn
Gửi thư

Group-invariant random processes

Luận án · Tiến sĩ · năm 2006 · đề tài tham khảo

Loại tài liệuLuận án
Bậc họcTiến sĩ
Tác giảAdam Timar
Cơ sở đào tạoĐại học Quốc gia TP. Hồ Chí Minh
Năm2006
NguồnThư viện số Đại học Quốc gia TP. Hồ Chí Minh · trang gốc

Tóm tắt

The dissertation centers around random processes on Rd or on transitive graphs that are invariant under the group of isometries or group of automorphisms, respectively. The first chapter is about the former, when we consider some isometry-invariant point process on Rd and examine what types of graphs can be defined on the points of the process in such a way that the point process and the graph has an equivariant distribution. One of particular interest, with applications in statistics, is when the defined graph is a biinfinite path. The rest of the dissertation is about random subgraphs of infinite connected transitive graphs (such as infinite Cayley graphs). We mostly consider so-called unimodular graphs, which is a natural and standard restriction in the field, and includes most of the known examples (all Cayley graphs, for example). One of the most-studied invariant random subgraph is Bernoulli percolation, when we delete each edge of the underlying graph with probability 1 - p, independently of each other. Chapters 2 and 3 consider questions aris¬ing in this setting, while Chapter 4 is about certain random subtrees, so-called minimal spanning forests, which are related to Bernoulli percolation and are also interesting on their own. Finally, in Chapter 5, we go beyond unimodular graphs, and generalize some results that have only been known for unimodular graphs to nommimodular graphs. We also show some remarkable differences between these two classes from the point of view of the behavior of invariant random subgraphs. In Chapter 1, I construct 1-ended tree factors and Zn factors for automorphism invariant point processes in 1".d that satisfy some nontriviality condition. (A factor is a measurable mapping of the point configuration to a graph on it that is equivariant with the point process.) The fact that such a construction exists for any d, independently of n, is perhaps surprising. These answer questions of Holroyd and Peres, and of Steve Evans. Given a transitive graph G, define p,, (G) := inf{p : there is a unique infinite cluster at Bernoulli(p) percolation}. The natural question whether every 1-ended transitive graph satisfies p G 1 was first asked by Benjamini and Schramm in their ground-breaking paper. Babson and Benjamini later showed that the answer is positive for the Cayley graphs of finitely presented groups. In Chapter 2, I give a simpler proof for their result and solve the three questions posed in their paper. These are of pure graph-theoretical nature, implicitly related to percolation. Two of these concern how certain properties of cutsets are preserved under quasi-isometries. Babson and Benjamini did not know whether the assumption "finitely presented" is necessary for their method in proving pu G 1 to work, or whether the key property that they actually need is satisfied by every 1-ended infinite Cayley graph. In this chapter I point out that the lamplighter group does not have this "key property", hence answering their third question. A transitive graph G is called unimodular if the left-invariant Haar measure on its group of automorphisms coincides with the right-invariant Haar measure. Every Cayley graph is unimodular, but not every transitive graph is. A natural question about the relation of infinite percolation clusters is the following, posed by Hiiggstrom, Peres and Schonmann. If there are infinitely many infinite clusters at Bernoulli percolation on a transitive graph, can then two of these clusters contain infinitely many vertices that are adjacent to the other cluster? I give a negative answer to this question for unimodular graphs in Chapter 3. Minimal spanning trees on finite graphs have played an important role in optimization for a while, but the interest towards their generalization for infinite graphs has arisen because of their close connection to percolation. I show in Chapter 4 that all the trees in the free minimal spanning forest of an infinite transitive unimodular graph have the
Bản toàn văn lưu tại thư viện Đại học Quốc gia TP. Hồ Chí Minh

Thông tin và tóm tắt ở trên lấy từ kho số mở của thư viện trường. Bản toàn văn do thư viện trường quản lý — có nơi cho đọc trực tuyến, có nơi cần tài khoản bạn đọc của trường.

Mở tại thư viện trường ↗
Dùng tài liệu tham khảo cho đúng
  • Đọc để hiểu cách người trước đặt câu hỏi, chọn phương pháp, trình bày số liệu — rồi tự làm với dữ liệu và bối cảnh của bạn.
  • Ý, số liệu, bảng biểu lấy từ tài liệu này phải trích dẫn (tác giả, năm, tên đề tài, cơ sở đào tạo).
  • Đề tài trùng tên vẫn làm được nếu khác đơn vị khảo sát, thời gian hoặc cách tiếp cận — nêu rõ điểm khác trong phần mở đầu.

Đề tài cùng ngành

Tác giả muốn gỡ, sửa thông tin hoặc bổ sung bản toàn văn: gửi yêu cầu qua trang Liên hệ, kèm đường dẫn trang này.

Tải ứng dụng AI Học Thuật

Tư vấn học thuật, luận văn, công bố quốc tế với chuyên gia — miễn phí trên iPhone, iPad và Android.

Download on the App Store Tải trên Google Play