Multilayer network analysis has become a vital tool for understanding different relationships and their interactions in complex system, where each layer multilayer depicts the topological structure of group nodes corresponding to particular relationship. The among layers imply how interplay relations on topology layer. For single-layer network, embedding methods have been proposed project into continuous vector space with relatively small number dimensions, embeds social representations...

Bayesian network (BN) has been adopted as the underlying model for representing and inferring uncertain knowledge. As basis of realistic applications centered on probabilistic inferences, learning a BN from data is critical subject machine learning, artificial intelligence, big paradigms. Currently, it necessary to extend classical methods BNs with respect data-intensive computing or in cloud environments. In this paper, we propose parallel incremental approach massive, distributed,...

Finding communities in multilayer networks is a vital step understanding the structure and dynamics of these layers, where each layer represents particular type relationship between nodes natural world. However, most community discovery methods for may ignore interplay layers or unique topological layer. Moreover, them can only detect non-overlapping communities. In this paper, we propose new method networks, which leverages topology to reveal overlapping Through comprehensive analysis edge...

A fast and efficient technique for profilometric measurement with a color-coded grating is proposed. Eight colors are used to code the grating, each color represents only one logical state. There 64 stripes in period of which large enough normal measurement. Compared previous techniques, it has advantages simple hardware without moving mechanical parts, single exposure obtaining three-dimensional information, little influence from noise nonlinearity CCD camera on accuracy, higher...

Inspired by the generation power of generative adversarial networks (GANs) in image domains, we introduce a novel hierarchical architecture for learning characteristic topological features from single arbitrary input graph via GANs. The consisting multiple GANs preserves both local and global automatically partitions into representative "stages" feature learning. stages facilitate reconstruction can be used as indicators importance associated structures. experiments show that our method...

The multiple relationships among objects in complex systems can be described well by multiplex networks, which contain rich information of the connections between objects. null model used to quantify specific nature a network, is powerful tool for analysing structural characteristics systems. However, networks remains largely unexplored. In this paper, we propose based on node redundancy degree, natural measure describing networks. Based model, define modularity study community structures...

This brief presents a general framework for the continuous-time nonlinear event-based state estimation problem. Using information from observations made by sampling, goal of problem is to estimate stochastic differential equations which represent uncertain system dynamics. challenging because measurements are taken only if some events happen rather than with fixed sampling interval. In this brief, theoretical solution derived and numerical algorithm based on Markov chain approximation...

The EmoInt-2017 task aims to determine a continuous numerical value representing the intensity which an emotion is expressed in tweet. Compared classification tasks that identify 1 among n emotions for tweet, present can provide more fine-grained (real-valued) sentiment analysis. This paper presents system uses bi-directional LSTM-CNN model complete competition task. Combining LSTM and CNN, prediction process considers both global information tweet local important information. proposed...

As mobile customers gradually occupying the largest share of cloud service users, effective and cost-sensitive provisioning services quickly becomes a main theme in computing. The key issues involved are much more than just enabling users to access remote resources through wireless networks. resource limited intermittent disconnection problems environments have intrinsic conflict with continuous connection assumption usage patterns. We advocate that seamless can only be achieved full...

Most classical search engines choose and rank advertisements (ads) based on their click-through rates (CTRs). To predict an ad’s CTR, historical click information is frequently concerned. accurately the CTR of new ads challenging critical for real world applications, since we do not have plentiful data about these ads. Adopting Bayesian network (BN) as effective framework representing inferring dependencies uncertainties among variables, in this paper, establish a BN-based model to CTRs...

Bayesian network (BN) with latent variables (LVs) provides a concise and straightforward framework for representing inferring uncertain knowledge unobservable or regard to missing data. To learn the BN LVs consistently realistic situations, we propose information theory based concept of existence weight incorporate it into clique-based learning method. In line challenges when LVs, focus on determining number relationships between observed variables. First, define algorithms finding ε-cliques...

Social networks can be modeled by graphs with nodes and edges, communities are sub within - groups of which connections dense, but between them sparser. According to this property communities, paper proposes a new modularity for measuring how good particular division is based on the concept coupling coefficient. Further more, applies proposed synthetic network data compares computational results under different modularity. The experimental show that our suitable cases all have nearly same...

Interval data are widely used in real applications to represent the values of quantities uncertain situations. However, implied probabilistic causal relationships among interval-valued variables with interval cannot be represented and inferred by general Bayesian networks point-based probability parameters. Thus, it is desired extend network effective mechanisms representation, learning inference data. In this paper, we define probabilities, bound-limited weak conditional probabilities...