检索结果-南通市图书馆

Distributed Aggregation Algorithms for Mobile Sensor Networks with Group Mobility Model

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2012年第5期17卷 512-520页

作者： Qianqian Ren Longjiang Guo jinghua zhu Meirui Ren Junqing zhu School of Computer Science and Technology Heilongjiang UniversityHarbin 150080China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin 150080China

In many applications of mobile sensor networks, such as water flow monitoring and disaster rescue, the nodes in the network can move together or separate temporarily. The dynamic network topology makes traditional spanning-tree-based aggregation algorithms invalid in mobile sensor networks. In this paper, we first present a distributed clustering algorithm which divides mobile sensor nodes into several groups, and then propose two distributed aggregation algorithms, Distance-AGG （Aggregation based on Distance）, and Probability-AGG （Aggregation based on Probability）. Both of these two algorithms conduct an aggregation query in three phases： query dissemination, intra-group aggregation, and inter-group aggregation. These two algorithms are efficient especially in mobile networks. We evaluate the performance of the proposed algorithms in terms of aggregation accuracy, energy efficiency, and query delay through ns-2 simulations. The results show that Distance-AGG and Probability-AGG can obtain higher accuracy with lower transmission and query delay than the existing aggregation algorithms.

关键词： mobile sensor networks data aggregation group mobility model distributed algorithms

Attention Neural Network for User Behavior Modeling

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2019年第1期 40-41页

作者： Kang Yang jinghua zhu School of Computer Science and Technology Heilongjiang University Harbin 150080 China

The recommendation system can effectively and quickly provide valuable information for users by filtering out massive useless data. User behavior modeling can extract all kinds of aggregated features over the heterogeneous behaviors to help recommendation. However, the existing user behavior modeling method cannot solve the cold-start problem caused by data sparse. Recent recommender systems which exploit reviews for learning representation can alleviate the above problem to a certain extent. Therefore, a user behavior modeling is proposed for recommendation task using attention neural network based on user reviews (AT-UBM). Firstly vanilla attention was used to sample reviews, and then CNN+Pooling method was applied to extract user behavior features. Finally the long-term behavior was combined with short-term behavior in feature spaces. Experimental results on real datasets show that the review-based user behavior model has better prediction accuracy and generalization capability.

关键词： Recommendation system Attention neural network Behavior model CNN+Pooling

CP-Net:Channel Attention and Pixel Attention Network for Single Image Dehazing

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2020年第1期 577-590页

作者： Shunan Gao jinghua zhu Yan Yang School of Computer Science and Technology Heilongjiang UniversityHarbin 150080China

An end-to-end channel attention and pixel attention network(CP-Net)is proposed to produce dehazed image directly in the *** CP-Net structure contains three critical ***,the double attention(DA)module consisting of channel attention(CA)and pixel attention(PA).Different channel features contain different levels of important information,and CA can give more weight to relevant information,so the network can learn more useful ***,haze is unevenly distributed on different pixels,and PA is able to filter out haze with varying weights for different *** sums the outputs of the two attention modules to improve further feature representation which contributes to better dehazing ***,local residual learning and DA module constitute another important component,namely basic block *** residual learning can transfer the feature information in the shallow part of the network to the deep part of the network through multiple local residual connections and enhance the expressive ability of ***,CP-Net mainly uses its core component,DA module,to automatically assign different weights to different features to achieve satisfactory dehazing *** experiment results on synthetic datasets and real hazy images indicate that many state-of-the-art single image dehazing methods have been surpassed by the CP-Net both quantitatively and qualitatively.

关键词： Image dehazing Channel attention and pixel attention Residual learning

Incremental Influence Maximization for Dynamic Social Networks

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2017年第2期 4-5页

作者： Yake Wang jinghua zhu Qian Ming School of Computer Science and Technology Heilongjiang University Harbin 150001 China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin China

Influence maximization is one fundamental and important problem to identify a set of most influential individuals to develop effective viral marketing strategies in social network. Most existing studies mainly focus on designing efficient algorithms or heuristics to find Top-K influential individuals for static network. However, when the network is evolving over time, the static algorithms have to be re-executed which will incur tremendous execution time. In this paper, an incremental algorithm DIM is proposed which can efficiently identify the Top-K influential individuals in dynamic social network based on the previous information instead of calculating from scratch. DIM is designed for Linear Threshold Model and it consists of two phases: initial seeding and seeds updating. In order to further reduce the running time, two pruning strategies are designed for the seeds updating phase. We carried out extensive experiments on real dynamic social network and the experimental results demonstrate that our algorithms could achieve good performance in terms of influence spread and significantly outperform those traditional static algorithms with respect to running time.

关键词： Influence maximization Dynamic social network Linear threshold model Pruning strategy

Target Market Optimal Coverage Algorithm Based on Heat Diffusion Model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2019年第2期 508-510页

作者： jinghua zhu Yuekai Zhang Bochong Li School of Computer Science and Technology Heilongjiang University Harbin 150080 China

The maximization of personalized influence is a branch of maximizing the influence of social networks, and the goal is to target specific social network users and mine the set of initial impact diffusion users that have made the most impact. However, most of the existing methods are based on the IC model and the LT model. The prediction of the impact of these two models on the nodes depends on the Monte Carlo simulation. In order to avoid Monte Carlo simulation time and simulate real life more, this paper introduces the heat diffusion model into the problem of maximizing the influence of personalization. The heat diffusion process was used to simulate the diffusion process of information influence. And the thermal energy was applied to measure the impact on the target users, and cluster candidate users. The cluster center as a seed node was proposed to spread information and maximizing the impact on specific users. The comparison experiments on real social networks show that the personalized maximization algorithm based on the thermal diffusion model has better time performance and diffusion effect than the traditional diffusion model.

关键词： Social network Influence maximization Heat diffusion Clustering

Image Segmentation of Cerebrospinal Fluid Using a New Hybrid Level-Set Model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Image Segmentation of Cerebrospinal Fluid Using a New Hybrid...

POI Recommendations Using Self-attention Based on Side Information

2011 International Conference on Computer Science and Information Technology(ICCSIT 2011)

作者： jinghua zhu Ying Ju Computer Science Department Xiamen University

ISBN: (纸本)9781612848334

It is very significant for medical diagnosis to get the correct boundary of cerebrospinal fluid (CSF) space in brain Magnetic Resonance (MR) images. In this paper, a new hybrid active contour model in the level-set framework for medical image segmentation was presented. The model is the hybrid of boundary-based and region-based. The boundaries of target objects can be extracted with the model straightforwardly requiring no preprocessing. The experimental results show that the proposed method is efficient.

关键词： CSF MRI active contour model level-set image segmentation

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2020年第2期 62-76页

作者： Chenbo Yue jinghua zhu Shuo Zhang Xinxing Ma School of Computer Science and Technology Heilongjiang UniversityHarbin 150080China

Point of interest(POI)recommendation is one of the most important tasks in location-based social networks(LBSN).The existing recommendation methods face two challenges:(1)the cold start problem caused by data sparsity;(2)underutilization of the abundant side information besides user-POI interaction in large-scale *** research shows that a user’s social relationship can be used to solve the cold start problem to some *** deep neural network learns users’long term and short term preferences to improve the recommendation ***,this paper proposes a POI recommendation model called SSANet,applying side information(S)and self-attention(SA)to provide the high-satisfaction POI recommendations for ***,first,the user-POI interaction matrix were constructed by users history data to represents the user hidden representation;second,the side information includes rating scores,access frequency,social relationship,and geographic information were used to extract users preference;third,we use self-attention mechanism to learn user long term and short term *** experimental results on the real LBSN datasets show that the recommendation performance of the SSANet model is better than the existing POI recommendation model.

关键词： Location-based social network Smart computing Point of interest Location-based services and applications

Multi-factor Fusion POI Recommendation Model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2020年第2期 21-35页

作者： Xinxing Ma jinghua zhu Shuo Zhang Yingli Zhong School of Computer Science and Technology Heilongjiang UniversityHarbin 150080China

In the context of the rapid development of location-based socialnetworks (LBSN), point of interest (POI) recommendation becomes an importantservice in LBSN. The POI recommendation service aims to recommendsome new places that may be of interest to users, help users to better understandtheir cities, and improve users’ experience of the platform. Although the geographicinfluence, similarity of POIs, and user check-ins information have beenused in the existing work recommended by POI, little existing work consideredcombing the aforementioned information. In this paper, we propose to makerecommendations by combing user ratings with the above information. Wemodel four types of information under a unified POI recommendation frameworkand this model is called extended user preference model based on matrixfactorization, referred to as UPEMF. Experiments were conducted on two realworld datasets, and the results show that the proposed method improves theaccuracy of POI recommendations compared to other recent methods.

关键词： Multi-factor fusion model Matrix factorization Euclidean distance Personalized recommendation

Interest-Forgetting Markov Model for Next-Basket Recommendation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

国际计算机前沿大会会议论文集 2019年第1期 29-31页

作者： jinghua zhu Xinxing Ma Chenbo Yue Chao Wang School of Computer Science and Technology Heilongjiang University Harbin 150080 Heilongjiang China

Recommendation systems provide users with ranked items based on individual’s preferences. Two types of preferences are commonly used to generate ranking lists: long-term preferences which are relatively stable and short-term preferences which are constantly changeable. But short-term preferences have an important real-time impact on individual’s current preferences. In order to predict personalized sequential patterns, the long-term user preferences and the short-term variations in preference need to be jointly considered for both personalization and sequential transitions. In this paper, a IFNR model is proposed to leverage long-term and short-term preferences for Next-Basket recommendation. In IFNR, similarity was used to represent long-term preferences. Personalized Markov model was exploited to mine short-term preferences based on individual’s behavior sequences. Personalized Markov transition matrix is generally very sparse, and thus it integrated Interest-Forgetting attribute, social trust relation and item similarity into personalized Markov model. Experimental results are on two real data sets, and show that this approach can improve the quality of recommendations compared with the existed methods.

关键词： Markov Social trust Next-Basket Recommendation Interest-Forgetting