检索结果-内蒙古大学图书馆

International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery, CyberC

作者： Chao Fan Junxuan Huang Dan Yang Zhihai Rong College of Arts and Sciences Shanxi Agricultural University Taigu China CompleX Lab University of Electronic Science and Technology of China Chengdu China Big Data Research Center University of Electronic Science and Technology of China Chengdu China

ISBN: (纸本)9781509051557

Point-of-interest (POI) denotes the category of a location people visit, reflecting the daily life and personal interests. Through a massive dataset of check-in records, this paper discloses that the POI visiting pattern of human spatial mobility falling into both Zipf's law and Heaps' law. It is found that people always have three potential choices when selecting a particular POI as their next travel destination, namely, exploring a new POI, returning to a visited one, or staying at the current one. We calculate the probabilities of these scenarios and propose a statistical model based on random walk to represent such phenomena. Simulation results show that our model can well reproduce the heterogeneous Zipf's law and Heaps' law of human mobility.

关键词： Simulation Companies Pragmatics Time-frequency analysis Gaussian distribution Distributed computing

来源：评论

学校读者我要写书评

暂无评论

Analyzing information sharing strategies of users in online social networks

Analyzing information sharing strategies of users in online ...

引用

International Conference on Advances in Social Network Analysis and Mining, ASONAM

作者： Dong-Anh Nguyen Shulong Tan Ram Ramanathan Xifeng Yan Department of Computer Science University of California Santa Barbara CA USA Baidu Big Data Lab Sunnyvale CA USA Network Research Department Raytheon BBN Technologies Cambridge MA USA

ISBN: (纸本)9781509028474

User information sharing is an important behavior in online social networks. Understanding such behavior could help in various applications such as user modeling, information cascade analysis, viral marketing, etc. In this paper, we aim to understand the strategies users employ to make retweet decision. We are interested in investigating whether these strategies in online social network contain significant information about users and can be used to further characterize users. We propose a flexible model that captures a number of behavior signals affecting user's retweet decision. Our empirical results show that the inferred strategies can help increase the performance of retweet prediction.

关键词： Information management Twitter Predictive models Pragmatics Mathematical model Electronic mail

来源：评论

学校读者我要写书评

暂无评论

IABC robotic evolutionary model for the foreign exchange rate prediction in Central America trading agreement events

IABC robotic evolutionary model for the foreign exchange rat...

引用

IEEE Symposium Series on Computational Intelligence (SSCI)

作者： Pei-Wei Tsai Jing Zhang Yao He Jui-Fang Chang Li-Hui Yang Wein-Duo Yang College of Information Science and Engineering Fujian Provincial Key Lab. of Big Data Mining & Apps. Fujian University of Technology Fuzhou China Department of International Business National Kaohsiung University of Applied Science Kaohsiung Taiwan

ISBN: (纸本)9781509042418

Taiwanese economy is extremely export-oriented. However, Taiwan also starts to actively look for opportunities for participation, because of the slowdown of multinational liberalizations, and as a result of the advance of the regional economic integration trend. As an example, Taiwan vigorously participates in FTAs with countries in Asia and Central America. International economic conditions are significantly important to the international trade and the volume of export/import volume, which is affected by the foreign exchange rate. If contemporary researchers could take full advantage of the exchange rate forecasting, Taiwan could maximize its trade surplus, thus boosting the economic growth. Conventional foreign exchange rate forecasting is usually provided by analyzing many financial indices or with the time-series method. Our goal is to produce the foreign exchange rate forecasting result by the robotic way with an evolutionary computing method called Interactive Artificial Bee Colony algorithm. Based on the event study methodology, the selected agreements include four FTA that are ECFA, BIA, ASTEP and ANZTEC, and the observation period setting is 70 days of pre-event period and 70 days of post-event period. This paper uses time series model (GARCH, EGARCH) and Interactive artificial bee colony (IABC) to establish the exchange rate predicting models. In addition, we adopt Mean Absolutely Percentage Error (MAPE) to compare the accuracy of exchange rate prediction. There are many exchange rate predicting models and the most frequently one to conduct maybe the time series model. This research reveals that even the IABC is relatively new it is the model has the best predictive ability among all the models.

关键词： Exchange rates Predictive models Forecasting Mathematical model Time series analysis

来源：评论

学校读者我要写书评

暂无评论

Frameworks to encode user preferences for inferring topic-sensitive information networks

Frameworks to encode user preferences for inferring topic-se...

引用

SIAM International Conference on data Mining 2015, SDM 2015

作者： Hu, Qingbo Xie, Sihong Lin, Shuyang Fan, Wei Yu, Philip S. Department of Computer Science University of Illinois at Chicago Chicago United States Baidu Research Big Data Lab. Sunnyvale United States Institute for Data Science Tsinghua University Beijing China

ISBN: (纸本)9781510811522

The connection between online users is the key to the success of many important applications, such as viral marketing. In reality, we often easily observe the time when each user in the network receives a message, yet the users' connections that empower the message diffusion remain hidden. Therefore, given the traces of disseminated messages, recent research has extensively studied approaches to uncover the underlying diffusion network. Since topic related information could assist the network inference, previous methods incorporated either users' preferences over topics or the topic distributions of cascading messages. However, methods combining both of them may lead to more accurate results, because they consider a more comprehensive range of available information. In this paper, we investigate this possibility by exploring two principled methods: Weighted Topic Cascade (WTC) and Preference-enhanced Topic Cascade (PTC). WTC and PTC formulate the network inference task as non-smooth convex optimization problems and adopt coordinate proximal gradient descent to solve them. Based on synthetic and real datasets, substantial experiments demonstrate that although WTC is better than several previous approaches in most cases, it is less stable than PTC, which constantly outperforms other baselines with an improvement of 4%∼10% in terms of the F-measure of inferred networks. Copyright © SIAM.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Image reconstruction algorithm of compressed sensing based on neural dynamics

引用

Journal of Information Hiding and Multimedia Signal Processing 2017年第3期8卷 623-631页

作者： Gan, Zhenhua Xiong, Baoping Zou, Fumin Liao, Lyuchao Du, Min College of Information Science and Engineering Fujian University of Technology No.3 Xueyuan Road University Town Minhou Fuzhou CityFujian Province China College of Electrical Engineering and Automation Fuzhou University No.2 Xueyuan Road University Town Minhou Fuzhou CityFujian Province China Fujian Provincial Key Laboratory of Big Data Mining and Applications No.3 Xueyuan Road University Town Minhou Fuzhou CityFujian Province China Key Lab of Automotive Electronics and Electric Drive Technology of Fujian Province No.3 Xueyuan Road University Town Minhou Fuzhou CityFujian Province China

To improve the recovery performance of compressed sensing (CS) recon- struction, this paper proposed a CS reconstruction model based on neural dynamics optimization algorithm (NDOA) with l1-norm minimization, and applied it to the two- dimensional (2D) image signals to verify the reconstruction effectiveness. Experiments proved the NDOA had performed better in reconstruction than the orthogonal matching pursuit (OMP) and its improved algorithms considering its perfect recovery. When the compressed sampling rate is 0:875 for 64×64 Lena image, the application of NDOA would raise the peak signal to noise ratio (PSNR) by 3:6466dB while decrease the relative error (REOR) by 2:21% in comparison with OMP and SWOMP. It showed that the effective- ness of CS reconstruction of NDOA was evident. At the same time, the adoption of NDOA could realize real-time processing by parallel computing, which would effectively solve real-time problems such as image processing in the realm of high dimensional sparse signals. © 2017.

关键词： Compressed sensing

来源：评论

学校读者我要写书评

暂无评论

Knowledge graph embedding translation based on constraints

引用

Journal of Information Hiding and Multimedia Signal Processing 2017年第5期8卷 1119-1131页

作者： Wu, Yunbing Pan, Jengshyang Lu, Ping Lin, Kaibiao Yu, Xiaoyan College of Mathematics and Computer Science Fuzhou University No.2. Xueyuan Road University Town Minhou Fuzhou CityFujian Province350116 China Fujian Provincial Key Lab of Big Data Mining and Applications Fujian University of Technology No.3. Xueyuan Road University Town Minhou Fuzhou CityFujian Province350118 China Department of Business Xiamen University of Technology No.600 Ligong Road Jimei District XiamenFujian Province361024 China School of Computer and Information Engineering Xiamen University of Technology No.600 Ligong Road Jimei District XiamenFujian Province361024 China

Knowledge graph reasoning is discovering new entity relations by computing and inference from existing relations. However, most reasoning models of translation embedding-based knowledge graphs have not considered the semantic-type constraints of relations in the construction of corrupted triplets. Hence, the constructed corrupted triplets may not conform to the actual semantic information and may, thus, significantly affect the prediction accuracy of the model. Therefore, we propose a constraint-based embedding model in this paper. First, the model establishes the head and tail entity set for each relationship. Then, it ensures that both the replacing head and tail entities in the corrupted triplet belong to the respective entity set so that the corrupted triplets that do not conform to the responding semantic relations are excluded. To evaluate the proposed model, we conduct link prediction and triple classification on WordNet and Freebase databases. The experimental results show that our method remarkably improves the performance compared to several state-of-the-art baselines. © 2017, Ubiquitous International. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Constraint Free Preference Preserving Hashing for Fast Recommendation

Constraint Free Preference Preserving Hashing for Fast Recom...

引用

IEEE Global Communications Conference

作者： Yan Zhang Guowu Yang Defu Lian Hong Wen Jinsong Wu Center for Cyber Security University of Electronic Science and Technology of China Big Data Research Center University of Electronic Science and Technology of China National Key Lab of Commun. University of Electronic Science and Technology of China Department of Electrical Engineering Universidad de Chile

ISBN: (纸本)9781509013296

Recommender systems have been widely used to deal with information overload, by suggesting relevant items that match users' personal interest. One of the most popular recommendation techniques is matrix factorization (MF). The inner products of learned latent factors between users and items can estimate users' preferences for items with high accuracy, but the preferences ranking is time consuming. Thus, hashing-based fast search technologies were exploited in recommender systems. However, most previous approaches consist of two stages: continuous latent factor learning and binary quantization, but they didn't well deal with the change of inner product arising from quantization. To this end, in this paper, we propose a constraint free preference preserving hashing method, which quantizes both norm and similarity in dot product. We also design an algorithm to optimize the bit length for norm quantization. The performance of our method is evaluated on three real world datasets. The results confirm that the proposed model can improve recommendation performance by 11%-15%, as compared with the state-of-the-art hashing approaches.

关键词： Recommender system Preference ranking Nor-m quantization Optimal bit Recommender systems Dot product Quantization hashing Preferences User preferences preference Consumers

来源：评论

学校读者我要写书评

暂无评论

Efficient algorithms to discover flock patterns in trajectories 16

Efficient algorithms to discover flock patterns in trajector...

引用

16th Brazilian Symposium on GeoInformatics, GEOINFO 2015

作者： Tanaka, Pedro Sena Vieira, Marcos R. Kaster, Daniel S. Department of Computer Science University of Londrina Brazil Big Data Lab Hitachi America Ltd. - R and D Brazil

The wide and increasing availability of GPS and location tracking devices has led to research advances in discovering spatiotemporal pattern in very large trajectory datasets. The overall objective of such patterns is to discover how spatial relationships between moving objects (e.g., vehicles, people, animals) behave over time. Among such patterns, one interesting and important category is flock pattern. This pattern is defined as a set of moving objects with minimum size that stay together within a maximum distance for a continuous period of time. Typical application examples are monitoring and surveillance, where they both rely on efficiently identifying groups of suspicious people/vehicles in large spatiotemporal streaming data. Previous works proposed polynomial-time algorithms to the flock pattern problem with fixed time duration. In this paper, we improve on previous work by applying a plane sweeping technique and inverted indexes. Plane sweeping technique speeds up the detection of candidate groups in a particular timestamp, while inverted indexes are employed to quickly compare candidate disks across timestamps. Using a variety of large real-world trajectory datasets we show that our proposed methods are efficient compared to state-of-the-art solution. In our experiments we show our proposed methods are up to 46x faster than previous solution. © 2015 National Institute for Space Research INPE All rights reserved.

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

Contourlet and nearest feature line based feature extraction approach for one prototype sample problem

引用

Journal of Information Hiding and Multimedia Signal Processing 2016年第5期7卷 1052-1059页

作者： Pan, Jeng-Shyang Yan, Lijun Chu, Shu-Chuan Roddick, John F. Fujian University of International Studies and Trade Fujian Province China Fujian Provincial Key Lab of Big Data Mining and Applications Fujian University of Technology China Shenzhen Institute of Information Technology Shenzhen China School of Computer Science Engineering and Mathematics Flinders University Australia

In this paper, a novel algorithm for face recognition with one sample per person is proposed. The proposed algorithm is based on contourlet transformation. For simple prototype sample problem, many discriminant learning methods can not work. Because for most discriminant learning methods, the within class scatter of the prototype samples are very important. However, simple prototype sample problem does not have within class scatter. To enhance the representative capability of the prototype samples set, some new samples are generated using contourlet transformation. Multiple prototype samples for each class are constructed through the decomposition and reconstruction of original training images by contourlet transformation. Thus neighborhood discriminant nearest feature line analysis can be performed on the new database. The experimental results demonstrate the efficiency of the proposed algorithm. © 2016, Ubiquitous International. All Rights Reserved.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Comparative study of Middle East respiratory syndrome coronavirus using bioinformatics techniques

Comparative study of Middle East respiratory syndrome corona...

引用

IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2015

作者： Ahn, Insung Jang, Jin-Hwa Biomedical Prediction Technology Lab. Korea Institute of Science and Technology Information Daejeon Korea Republic of Department of Big Data Science University of Science and Technology Daejeon Korea Republic of

ISBN: (纸本)9781467367981

In this study, we collected 141 coding sequences of MERS-CoV from the National Center for Biotechnology and Information (NCBI), including the sequences isolated in Korea in 2015. We conducted the phylogenetic analysis using the Maximum Likelihood method to examine the overall variation patterns among the target sequences, and then, we divided the target sequences into 4 different groups according to occurred countries and host species. Using the codon analyzer named SimFluVar program, we analyzed the codon variation patterns in the wobble position among 4 groups. In order to investigate the effect of codon variations that can change the phenotype of target genes, we compared the transversional substitutions between the Korean-origin sequences and other groups. As a result, we found that the Korea-origin sequences showed very minor differences with those collected from the Saudi Arabia in 2015, whereas other groups which were collected from USA and UK in 2013 and 2014 showed more complicated differences. We also compared the Korea-origin sequences with those of camel-origin sequences, and we found that the substitution pattern was somewhat different with that of human-origin viruses. © 2015 IEEE.

关键词： Bioinformatics

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：