Algorithm 1.

Variational EM algorithm of GLBA

Input: A multi-graph ${\left\{I_{i,j}^k\in \{0,1\}\right\}}_{i,j\in {\mathrm{\Omega }}_{k^{\prime }}}0<\gamma <0.5$
Output: subject parameters $\mathrm{\Theta }=\left({\left\{\left({\tau }_i,{\alpha }_i,{\beta }_i\right)\right\}}_{i=1}^m,\gamma \right)$
	Initialisation : τ0 = 0.5, αi = βi = τi = 1.0, i = 1, . . . ,m
1:	repeat
2:	for k − 1 to n do
3:	compute statistics ${\tilde{\alpha }}_i^{(k)},{\tilde{\beta }}_i^{(k)},{\tilde{\tau }}_i^{(k)}$ by Eq. (18);
4:	end for
5:	for i − 1 to m do
6:	solve (αi, βi) from Eq. (19) (Newton-Raphson);
7:	compute τi by Eq. (20);
8:	end for
9:	(optional) update γ from Eq. (21);
10:	until ${\left\{\left({\tau }_i,{\alpha }_i,{\beta }_i\right)\right\}}_{i=1}^m$ are all converged.
11:	return Θ