Large Scale Community Discovery of Complex Networks Based On Approximate Optimization

Abstract

Study of structural features associated with complex networks has its utility. In the process a key role is played by community. Online networks are showing unprecedented growth in complexity. In this context, the traditional approach used to discovery community is not sufficient to deal with large networks. Thus it leads to the problem of inability to discovery communities efficiently in the large scale networks. To overcome this problem, a Parallel Community Discovery (PCD) algorithm is proposed based on the concept of approximate optimization. The algorithm has two parts. They are known as mountain model and landslide algorithm. The former is meant for providing approximate selection of nodes that are used for representation in graph theory. It shows users and the data posted by them in the network. The landslide algorithm on the other hand updates the modularity from time to time incrementally based on the aforementioned technique named approximate optimization. Thus the proposed methodology helps in discovering communities in parallel. The process of clustering on the modularity plays important role in mountain model as it employs weight associated with each edge in the given network. The communities also exhibit relationship among them. However, they are simplified by the landslide algorithm thus helping in extraction of community structural features from large networks.

A New Spectral Conjugate Gradient Method with Strong Wolfe-Powell Line Search

“…Several scaled CG methods proposed by many authors can be referred to [26][27][28][29][30][31][32][33][34][35], and an application problem can be referred to [44]. In this paper, the spectral conjugate gradient method SMMAR with strong Wolfe-Powell line search is presented as follows: In the next section, the new spectral conjugate gradient formula and the algorithm.…”

Design of an Aerial Drone Delivery System

“…Indoor delivery systems have even been developed. [18,19,20] Overall, multiple system implementation and optimization strategies are necessary, and efforts have been made to create such strategies [21,22,23].…”