From: Learning attribute and homophily measures through random walks
Notation | Description |
---|---|
\(\mathscr {P}(\alpha , \mu , f)\) | Model class \(\mathscr {P}\) (non-linear preferential attachment model with homophily) |
\({\mathcal {A}}\) | Attribute space |
\(\mu\) | Attribute distribution of an arriving node |
\(\alpha\) | Preferential attachment parameter |
\(f(a,a')\) | Propensity of a pair of nodes with attributes a and \(a'\) to interact |
m | Number of edges a new node entering the system connects to pre-existing nodes |
\(\deg _i(v,{n})\) | Degree of v at time n when i of the edges of \(v_{{n}+1}\) have connected to the network |
\(\mathscr {U}(\alpha , \nu , f)\) | Model class \(\mathscr {U}\) |
\(\nu\) | Resolvable measure |
\({\mathcal {E}}\) | Set of edges of the network |
N | Number of nodes in the network |
\({\mathcal {V}}_a\) | Set of nodes of type a |
\({\mathcal {E}}_{aa'}\) | Set of edges between nodes of attributes a and \(a'\) |
\(D_{a}\) | Dyadicity of nodes with attribute a |
\(H_{aa'}\) | Heterophilicity between nodes with attributes a and \(a'\) |
p(a) | Proportion of nodes with attribute a |
|.| | Number of elements of a set |
p(k|a) | proportion of nodes of degree k having attribute a |
\(\widehat{.}\) | Estimator of a quantity |
\(d_i\) | Degree of node i |
\(\pi _i\) | Probability of sampling node i |
\(\pi _{(i,j)}\) | Probability of sampling edge (i, j) |
\(w_{ij}\) | Eeight of edge (i, j) |
\(\theta\) | propensity of N2V to backtrack |
\(\gamma\) (\(\beta\)) | Propensity of a N2V to reach a (non-)common neighbor of the currently visited node |
and the previously visited node | |
\(\delta\) | Spectral gap |