本文選題：多層網(wǎng)絡(luò) + 社區(qū)發(fā)現(xiàn)��； 參考：《計(jì)算機(jī)與現(xiàn)代化》2017年06期

【摘要】：如何在多層網(wǎng)絡(luò)中發(fā)現(xiàn)社區(qū)是一項(xiàng)巨大挑戰(zhàn)。目前有些算法將多層網(wǎng)絡(luò)表示成三階張量,然后使用非負(fù)張量分解進(jìn)行社區(qū)發(fā)現(xiàn)。但在多層網(wǎng)絡(luò)的每層網(wǎng)絡(luò)中存在很多社區(qū)之間的連接或每層網(wǎng)絡(luò)都很稀疏的情況下,非負(fù)張量分解算法的準(zhǔn)確率較差。為了解決這一問題,本文提出一種改進(jìn)算法。先將原始多層網(wǎng)絡(luò)進(jìn)行層次約簡,減少多層網(wǎng)絡(luò)的層數(shù),使其社區(qū)結(jié)構(gòu)更加凸顯,然后再使用非負(fù)張量分解算法進(jìn)行社區(qū)發(fā)現(xiàn)。在人工數(shù)據(jù)集與真實(shí)數(shù)據(jù)集上的實(shí)驗(yàn)表明,本文所提出的框架在準(zhǔn)確率上有明顯的優(yōu)勢。
[Abstract]:How to find communities in multilayer networks is a huge challenge. At present, some algorithms represent multi-layer networks as third-order Zhang Liang, and then use non-negative Zhang Liang decomposition for community discovery. However, the accuracy of non-negative Zhang Liang decomposition algorithm is poor when there are many connections between communities in each layer of multilayer network or every layer network is sparse. In order to solve this problem, this paper proposes an improved algorithm. Firstly, the original multi-layer network is reduced to reduce the number of layers, so that the community structure is more prominent, and then the non-negative Zhang Liang decomposition algorithm is used for community discovery. Experiments on artificial data sets and real data sets show that the proposed framework has obvious advantages in accuracy.
【作者單位】： 北京交通大學(xué)計(jì)算機(jī)與信息技術(shù)學(xué)院交通數(shù)據(jù)分析與挖掘北京市重點(diǎn)實(shí)驗(yàn)室;
【基金】：國家自然科學(xué)基金資助項(xiàng)目(61403023) 教育部-中國移動科研基金資助項(xiàng)目(MCM20150513) 中國博士后科學(xué)基金資助項(xiàng)目(2015M580040)
【分類號】：TP301.6
，

本文編號：2002380