Spring 2025, Probability Topics: A Walk in the Random Park

Instructor: Jacob Richey

Course details:

Thursday 16:15-17:45, BME Building H Room H406. Neptun code BMETESZMsMVFVA-00

Suggested texts

Syllabus

This course will focus on three models: random walks, percolation, and first passage percolation.
For each topic, we will start by reviewing classical results and methods, and then move on to recent ideas and papers.

Papers for presentation

Here are the papers I suggest for presentations. Your presentation should consist roughly of two parts: describing the general structure of the paper or or of one of its theorems (~5-10 minutes), and a deep dive into a proof of one lemma or fact (~20-25 minutes).

Link to presentation topics schedule: link

Simple Random Walk

Percolation

First Passage Percolation

Schedule

Date Topics Relevant text Suggested exercises
Feb 13 Random walk basics: LLN, CLT, LCLT D Theorem 3.5.2 (HE) Exercises 1.8, 1.9; (RW) Exercises 1.4, 1.7, 2.7, 2.9
Feb 20 LCLT, Ornstein coupling, RW proof of Liouville theorem D Theorem 3.5.2 Show that two simple random walks in \(\mathbb{Z}^d\) can be coupled so that they are eventually equal. (Hint: use the Ornstein coupling idea in each dimension separately.) And: (RW2) Exercises 2.4, 2.6
Feb 27 Green's function (transient case) I Chapter 2, RW Chapters 4, 6 Let \(z = (1,0, \ldots, 0) \in \mathbb{Z}^d\). Show that the probability that a SRW \(S\) ever visits \(z\) is \(p = 1-G(0,0)^{-1}\). Let \(T\) be the first return time of \(S\) to the set \(\{0, z\}\). Show that \(\mathbb{P}(S_T = 0) = \mathbb{P}(S_T = z) = \frac{p}{1+p}\). And: (RW) Exercises 4.1, 4.6, 4.7
Mar 6 WLLN for Range of 2D SRW; Midpoint of 1D SRW Range is Uniform Erdos & Dvoretzky (1950), Sections 2-3, Discussion of Hot Potato puzzle Let \(M_n, m_n\) denote (respectively) the maximum/minimum of the range \(R_n\) of 1D SRW at time \(n\). Prove or disprove: \(\frac{M_n-m_n}{\sqrt{n}}\) converges in distribution as \(n \to \infty\). If it converges, identify the limit.
Mar 13 Last exit decomposition, capacity, Wiener's test, potential kernel in \(\mathbb{Z}^2\) I Chapter 2, RW Chapters 4, 6 Use the potential kernel in \(\mathbb{Z}^2\) to compute the asymptotic for the expected number of visits to site \(x\) by a simple random walk started at 0 and stopped on hitting 0 (again) as \(|x| \to \infty\).
Mar 20 Bernoulli Percolation in \(\mathbb{Z}^d\), Peirels argument DC Chapter 2 (DC) Exercises 5, 8, 10
Mar 27 Basic tools: Increasing coupling, FKG, Russo DC Chapter 2 (DC) Exercises 11, 14
April 3 \(p_c = \frac{1}{2}\) for Bernoulli bond percolation in \(\mathbb{Z}^2\), via influences Sharp threshold phenomena in statistical physics