arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1601.02658math.PRcond-mat.stat-mechcs.CCcs.ITcs.SImath.IT
收藏

稀疏网络中社团检测的信息论阈值

Information-theoretic thresholds for community detection in sparse networks

Jess Banks, Cristopher Moore

摘要:

我们给出了随机块模型中社区检测的信息论阈值的上界和下界。具体地,令$k$为组的数目,$d$为平均度,组内和组间顶点之间的边的概率分别为$c_\mathrm{in}/n$和$c_\mathrm{out}/n$,并且令$λ= (c_\mathrm{in}-c_\mathrm{out})/(kd)$。我们发现,当$k$是大的,$λ= O(1/k)$,社区检测成为可能的$d$的临界值-在物理方面,冷凝阈值-是\[ d_c = Θ\!\left( \frac{\log k}{k λ^2} \right) \, , \]与更严格的结果在某些制度。在此阈值以上,我们表明,只有分区的节点到$k$组与地面真相,给出一个指数时间算法,表现优于机会-特别是,检测是可能的$k \ge 5$在discretative情况下$λ< 0$和$k \ge 11$在discretative情况下$λ> 0$。(类似的上限由阿贝和桑登独立获得。在这个阈值以下,我们使用Neeman和Netrapalli(他们推广了Mossel,Neeman和Sly的论点)的最新结果来表明,没有算法可以比机会更好地标记顶点,甚至可以区分块模型和Erdens-Rényi随机图。我们还依赖于某些功能的双随机矩阵由于Achlioptas和Naor的界限;事实上,我们的下限$d_c$是二阶矩下界的$k$-着色阈值的随机图具有一定的有效程度。

英文摘要:

We give upper and lower bounds on the information-theoretic threshold for community detection in the stochastic block model. Specifically, let $k$ be the number of groups, $d$ be the average degree, the probability of edges between vertices within and between groups be $c_\mathrm{in}/n$ and $c_\mathrm{out}/n$ respectively, and let $λ= (c_\mathrm{in}-c_\mathrm{out})/(kd)$. We show that, when $k$ is large, and $λ= O(1/k)$, the critical value of $d$ at which community detection becomes possible -- in physical terms, the condensation threshold -- is \[ d_c = Θ\!\left( \frac{\log k}{k λ^2} \right) \, , \] with tighter results in certain regimes. Above this threshold, we show that the only partitions of the nodes into $k$ groups are correlated with the ground truth, giving an exponential-time algorithm that performs better than chance -- in particular, detection is possible for $k \ge 5$ in the disassortative case $λ< 0$ and for $k \ge 11$ in the assortative case $λ> 0$. (Similar upper bounds were obtained independently by Abbe and Sandon.) Below this threshold, we use recent results of Neeman and Netrapalli (who generalized arguments of Mossel, Neeman, and Sly) to show that no algorithm can label the vertices better than chance, or even distinguish the block model from an Erdős-Rényi random graph with high probability. We also rely on bounds on certain functions of doubly stochastic matrices due to Achlioptas and Naor; indeed, our lower bound on $d_c$ is the second moment lower bound on the $k$-colorability threshold for random graphs with a certain effective degree.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑