$$ \newcommand{\st}{\text{ s.t. }} \newcommand{\and}{\text{ and }} \DeclareMathOperator*{\argmin}{arg\,min} \DeclareMathOperator*{\argmax}{arg\,max} \DeclareMathOperator*{\liminf}{lim\,inf} \DeclareMathOperator*{\limsup}{lim\,sup} \DeclareMathOperator*{\dom}{dom} \DeclareMathOperator*{\epi}{epi} \newcommand{\<}{\langle} \newcommand{\>}{\rangle} \newcommand{\R}{\mathbb{R}} \newcommand{\N}{\mathbb{N}} \newcommand{\O}{\mathcal{O}} \newcommand{\dist}{\text{dist}} \newcommand{\vec}[1]{\mathbf{#1}} \newcommand{\diag}{\mathrm{diag}} \newcommand{\d}{\mathrm{d}} \newcommand{\L}{\mathcal{L}} \newcommand{\H}{\mathcal{H}} \newcommand{\Tr}{\mathrm{\mathbf{Tr}}} \newcommand{\E}{\mathbb{E}} \newcommand{\Var}{\mathrm{Var}} \newcommand{\Cov}{\mathrm{Cov}} \newcommand{\indep}{\perp \!\!\! \perp} \newcommand{\KL}[2]{\mathrm{KL}(#1 \parallel #2)} \newcommand{\W}{\mathbf{W}} % Wasserstein distance \newcommand{\SW}{\mathbf{SW}} % Sliced-Wasserstein distance $$

Page Rank & Community detection

Page Rank Let $G = (V, E)$ be a web graph. Each node $v \in V$ represents a web page, while there is a directed edge from $u \in V$ to $v \in V$ if $u$ has a hyperlink to $v$. The importance of $v$ is proportional to the importance of nodes linking to $v$. Let $\delta_\text{in}(v)$ and $\delta_\text{out}(v)$ be the in and out degrees of $v$ in $G$. Then we can define the matrix $M$ as follows: ...

September 13, 2024 · 5 min