PHOTO

Dr. PENG ZHANG

Associate Professor
Data Mining and Social Media Computing Group
Institute of Information Engineering, Chinese Academy of Science
91, Minzhuang Road, Haidian District, Beijing 100193, China
E-mail: zhangpeng@iie.ac.cn  
Phone: +861082546754 (Office)

 

 

Short Bio

Peng Zhang is an Associate Professor at the  Institute of Information Engineering, Chinese Academy of Sciences . His research includes data streams, social network analysis, and convex optimization for data classification. He received PhD in Computer Science from  Chinese Academy of Sciences in July 2009.

Research 1: Social Network Analysis

Social media, such as Facebook, Flickr, Twitter, have become important mediums, with rapidly increasing users over the past few years. Through the powerful effect of word-of-mouth, social media play a critical role in affecting people's opinions and behaviors. Social media analysis is an inherently interdisciplinary academic field emerged from social psychology, sociology, statistics and graph theory. Our research includes but not limited to: social influence maximization, social behavior analysis for targeting and advertising, Hadoop for distributed SQL querying on big social data, social media stream monitoring and topic modeling for social media content understanding. Our research is supported by the National Science Foundation of China (NSFC) and the Chinese Academy of Sciences Strategic Leading Project. Under the supports, we developed cutting-edge models and algorithms to better understand/manage social media for business/security applications.

  • (AAAI-14) Q. Zhi, P. Zhang, Y. Cao, C. Zhou, and L. Guo. Combining Heterogenous Social and Geographical Information for Event Recommendation, In Proceedings of the 28th AAAI Conference on Artificial Intelligence, July 27-31, 2014, Quebec City, Canada.
  • (ICCS-14) C. Zhou, P. Zhang, W. Zang, L. Guo. Maximizing the Cumulative Influence through a Social Network when Repeat Activation Exists. In Proceedings of the 14th International Conference on Computational Science, Cairns, Australia.
  • (ICCS-14) Y. Cao, P. Zhang, J. Guo, L. Guo. Mining Large-scale Event Knowledge from Web Text. In Proceedings of the 14th International Conference on Computational Science, Cairns, Australia, May 2014.
  • (ICCS-14) W. Zang, P. Zhang, C. Zhou, L. Guo. Discovering Multiple Diffusion Source Nodes in Social Networks. In Proceedings of the 14th International Conference on Computational Science, Cairns, Australia.
  • (WWW-14) C. Zhou, P. Zhang, W. Zang and L. Guo, Maximizing the Long-term Integral Influence in Social Networks under the Voter Model, In Proceedings of the 23rd ACM International World Wide Web Conference, Korea, 2014.
  • (WWW-14) C. Zhou, P. Zhang, J. Guo and L. Guo, Upper Bound based Greedy Algorithm for Mining Top-k Influential Nodes in Social Networks, In Proceedings of the 23rd ACM International World Wide Web Conference, Korea, 2014.
  • (CIKM-13) J. Guo, P. Zhang, C. Zhou, Y.Cao and L. Guo. Personalized Influence Maximization on Social Networks. In Proceedings of the 22nd International Conference on Information and Knowledge Management, October 28- November 1, 2013, San Francisco, CA, USA.  [MP3, PPT, PDF]
  • (ICDM-13) C. Zhou, P. Zhang, J. Guo, X. Zhu and L. Guo, UBLF: An upper bound based approach to discover influential nodes in  social networks. In Proceedings of the 13th IEEE International Conference on Data Mining, December 7-10, 2013, Dallas, TX, USA. [PDF , PPT]
  • (BigData-13) Y. Ma, P. Zhang, Y.Cao and L. Guo. Parallel Auto-encoder for Efficient Outlier Detection. In Proceedings of the 2013 IEEE International Conference on Big Data, October 6-9, 2013, Santa Clara, USA.

Research 2: Streaming and Online Learning

Data Stream represents a new class of data-intensive applications such as wireless sensor networks, web traffic management, telephone call records, on-line transactions, Internet bourse records, and web servers’ logs. As many modern information systems gradually evolve towards sophisticated data stream environments that can collect real time streaming data from thousands of terminals, the existing data management frameworks are hard to adapt to dynamic changing of the data streams, and are incapable of achieving fast response for real-time decision making.
Our research explored new theoretical foundations and technical solutions for building a stream-based data analytic framework for fast response and real-time decision making. By proposing solutions to attack key challenges at different levels, including data sampling, model training and prediction, and sub-linear time complexity for classification, our project eventually delivers a stream-based data mining framework to achieve adaptive sampling, future forecasting, and real-time classification for data stream applications.

  • (TKDE-14) P. Zhang, C. Zhou, P. Wang, B. Gao, X. Zhu, and L. Guo, E-Tree: An Efficient Indexing Structure for Ensemble Models on Data Streams. IEEE Transactions on Knowledge and Data Engineering, Vol. 26 (3), 2014.
  • (PAKDD-14) P. Wang, P. Zhang, L. Guo, Forward Prediction on Data Streams, In Proceedings of the 14th Pacific-Asia Conference on Knowledge Discovery and Data Mining , 2014.
  • (WWW-14) Z. Qiao, P. Zhang, J. He, Y. Cao, C. Zhou and L. Guo, Combining Geographical Information of Users and Content of Items for Accurate Rating Prediction, In Proceedings of the 23rd ACM International World Wide Web Conference, Korea, 2014.
  • (ICCS-14) Z. Qiao, P. Zhang, Y, Cao, C, Zhou, L, Guo. Improving Collaborative Recommendation via Location-based User-Item Subgroup. In Proceedings of the 14th International Conference on Computational Science, Cairns, Australia.
  • (ICCS-14) Z. Tan, P. Zhang, J. Tan, L. Guo. A Multi-layer Event Detection Algorithm for Detecting Global and Local Hot Events in Social Networks. In Proceedings of the 14th International Conference on Computational Science, Cairns, Australia.
  • (PAKDD-13) Z. Qiao, G. Huang, J. He, P. Zhang, L. Guo, J. Cao, and Y. Zhang, Discovering Semantics from Multiple Correlated Time Series Stream. In Proceedings of the 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining , Australia, April 14-17, 2013.
  • (Neurocomputing) P. Zhang, B. Gao, P. Liu, Y. Shi, and L. Guo, A Framework for Application-Driven Classification of Data Streams.  Neurocomputing, Volume 92, 2012, pages: 170-182.
  • (ICDM-12) J. Li, P. Zhang, Y. Cao, P. Liu, and L. Guo, Efficient Behavior Targeting Using Ensmeble SVM Indexing. In Proceedings of the 12th IEEE International Conference on Data Mining, Dec. 09-13, 2012, Brussels, Belgium. [PDF]
  • (SDM-12) P. Wang, P. Zhang, and L. Guo, Minining Multi-label Data Streams Using Ensemble-based Active Learning . In Proceedings of the 2012 SIAM International Conference on Data Mining, Anaheim, California, USA.
  • (ICCS-12) J. Guo, P. Zhang, J. Tan, L. Guo. Mining Hot Topics from Twitter Streams. In Proceedings of the 12th International Conference on Computational Science , Omaha, USA.
  • (APWeb-12) Z. Qiao, Jing He, Jie Cao, G. Huang, P. Zhang, Multiple Time Series Anomaly Detection Based on Compression and Correlation Analysis: A Medical Surveillance Case Study. In Proceedings of the 14th Asia-Pacific Web Conference, 294-305.
  • (DSS) P. Zhang, X. Zhu, Y. Shi, L. Guo, and X. Wu, Robust Ensemble Learning for Mining Noisy Data Streams, Decision Support Systems, Vol. 50(2), 2011, pages: 469-479.
  • (KDD-11) P. Zhang, J. Li, P. Wang, B. Gao, X. Zhu, and L. Guo, Enabling Fast Prediction for Ensemble Models on Data Streams. In Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , August 21-24, 2011, San Diego, CA, USA. [PDF, PPT]
  • (ICDM-11) P. Zhang, B. Gao, X. Zhu, and L. Guo, Enabling Fast Lazy Learning for Data Streams. In Proceedings of the 11th IEEE International Conference on Data Mining , December 11-14, 2011, Vancouver, Canada. [PDF]
  • (ICTAI-11) J. Li, P. Zhang, J. Tan,  L. Guo, Adaptive Shared-Filter Ordering for Efficient Multimedia Stream Monitoring", Proceedings of the 23rd IEEE International Conference on Tools with Artificial Intelligence ,Boca Raton, Florida, USA, 2011.
  • (CIKM-11) J. Guo, P. Zhang, J. Tan, and L. Guo, Mining Frequent Patterns across Multiple Data Streams. In Proceedings of the 20th ACM International Conference on Information and Knowledge Management , October 24-28, 2011, Glasgow, Scotland, UK.
  • (CIKM-11) J. Li, P. Zhang, J. Tan, P. Liu, and L. Guo, Continuous Data Stream Query in the Cloud. In Proceedings of the 20th ACM International Conference on Information and Knowledge Management , October 24-28, 2011, Glasgow, Scotland, UK.
  • (WI-11) J. Li, W. Zang, J. Tan, P. Zhang: Predictive Data Stream Filtering. Web Intelligence s 2011, pages: 237-240.
  • (TSMC-B) X. Zhu, P. Zhang, X. Lin, Y. Shi, Active Learning from Stream Data Using Optimal Weight Classifier Ensemble, IEEE Transactions on System, Man, Cybernetics, Part B, Vol. 40 (6),2010, pages: 1607-1621.
  • (ICDM-10) P. Zhang, X. Zhu, J. Tan, and L. Guo, Classifier and Cluster Ensembles for Mining Concept Drifting Data Streams, In Proceedings of the 10th IEEE International Conference on Data Mining , Sydney, Australia, December 14-17, 2010, pages: 1175-1180.
  • (CIKM-10) P. Zhang, X. Zhu, J. Tan, and L. Guo, SKIF: A Data Imputation Framework for Concept Drifting Data Streams, In Proceedings of the 19th ACM International Conference on Information and Knowledge Management , Toronto, Ontario, Canada, October26-31, pages: 1869-1872.
  • (ICCAS-10) Jun Li, Peng Zhang, Jianlong Tan, Li Guo, and Ping Liu, Make filters smart in multimedia streams environments", Proceedings of 2010 International Conference on Communications, Circuits and Systems.
  • (ICMLA-10) Z. Qiao, P. Zhang, J. He, J. Yan, L. Guo: Learning from Multiple Related Data Streams with Asynchronous Flowing Speeds. ICMLA 2010: 272-277
  • (ICDM-09) P. Zhang, X. Zhu, and L. Guo, Mining Data Streams with Labeled and Unlabeled Training Examples. In Proceedings of IEEE International Conference on Data Mining , Miami, Florida, USA, December 6-9, 2009, pages: 627-636.
  • (PAKDD-09) P. Zhang, X. Zhu, Y. Shi, and X. Wu, An Aggregate Ensemble for Mining Concept Drifting Data Streams with Noise. In Proceedings of the 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, Bangkok, Thailand, April 27-30, 2009, pages: 1021-1029.
  • (ICCS-09) P. Zhang, X. Zhu, and Y. Shi. Bias-Variance Analysis for Ensembling Regularized Multiple Criteria Linear Programming, International Conference on Computational Science , Baton Rouge, Louisiana, USA, May 25-27, 2009.
  • (KDD-08) P. Zhang, X. Zhu, and Y. Shi, Categorizing and Mining Concept Drifting Data Streams. In Proceedings of the 14th ACM international conference on Knowledge Discovery and Data mining, Las Vegas, Nevada, USA, August 24-27, 2008, pages: 812-820.
  • (ICDM-08) X. Zhu, P. Zhang, X. Wu, D. He, C. Zhang, and Y. Shi, Cleansing Noisy Data Streams. In Proceedings of the 8th IEEE International Conference on Data Mining , Pisa, Italy, December 15-19, 2008, pages: 1139-1144.
  • (ICDM-07) X. Zhu, P. Zhang, X. Lin and Y. Shi, Active Learning from Data Streams, In Proceedings of the 7th IEEE International Conference on Data Mining , Omaha, Nebraska, USA, October 28-31, 2007, pages: 757-762.

    Research 3: Convex Optimization for Data Classification

    For the last decade, researchers have extensively studied a quadratic program, known as Vapnik's Support Vector Machine (SVM) , for data classification. Using optimization techniques to classify data can trace back to 1960s when Kendall discussed an LP-related convex-hull procedure for classifying data groups. Then, Mangasarian and his group formulated linear program as a large margin classifier. In 1970s, Charnes and Cooper initiated Data Envelopment Analysis where a fractional programming is used to evaluate decision making units. In 1980s, Glover proposed a number of linear programming models to solve data discriminant problems with small training samples. In 1990s, the linear programming models were extended to multiple criteria linear programming (MCLP) for classification.  However, the structure of the MCLP cannot always guarantee a stable solution.
    Based on the convex optimization theory, we extended MCLP into strict convex optimization problems by adding different regularized terms to the objective function. Comparing to the original linear programming model, the strict convex formulations guarantee the unique solution and is always mathematically solvable. These new models have been successful applied to credit card portfolio management, bioinformatics, fraud management, intrusion detection and firm bankruptcy analysis.

  • (JOTB) R. Chen, Z. Zhang, D. Wu, P. Zhang, X. Zhang, Y. Wang, and Y. Shi, Prediction of protein interaction hot spots using rough set-based multiple criteria linear programming, Journal of Theoretical Biology,Vol. 269 (1), January 2011, Pages 174-180.
  • (WIAS) P. Zhang, X. Zhu, Z. Zhang, and Y. Shi, Multiple Criteria Programming for VIP Email Behavior Analysis. Web Intelligence and Agent Systems: An International Journal, Vol. 8(1), 2010, pages:  69-78.
  • (COR) J. Zhang, Y. Shi, and P. Zhang, Several Multi-criteria Programming Methods for Classification. Computers and Operations Research , Vol. 36(3), 2009, pages: 823-836.  
  • (IJDWM) D. Sun, L. Liu, P. Zhang, X. Zhu, Y. Shi, Decision Rule Extraction for Regularized Multiple Criteria Linear Programming Model,  International Journal of Data Warehousing and Mining.
  • (China Science) Y. Shi, Y. Tian, X. Chen, and P. Zhang, Regularized Multiple Criteria Linear Programs for Classification. Science in China Series F: Information Sciences, Vol. 52(10), 2009, pages: 1812-1820.
  • (IJOQM) P. Zhang, Y. Tian, Z. Zhang, X. Li, and Y. Shi. Supportive instances for Regularized multiple criteria linear programming, International Journal of Operations \& Quantitative Management, Vol. 4(4), 2008, pages: 249-263.
  • (IJOQM) Z. Zhang, P. Zhang, and Y. Shi. A Rough Set-Based Multiple Criteria Linear Programming Approach for Classification, International Journal of Operations \& Quantitative Management. Vol. 14(4), 2008.
  • (ICCS-10) Y. Zhang, P. Zhang, L. Zhang, and Y. Shi. Knowledge Extraction from Multiple Criteria Linear Programming Classification Approach, the 10th International Conference on Computational Science, Netherlands, Amsterdam, May 31-June 2, 2010.               
  • (ICDM-09) X. Zhou, W. Jiang, Y. Tian, P. Zhang, G. Nie, and Y. Shi, A New Kernel-based Classification Algorithm. In Proceedings of IEEE International Conference on Data Mining , Miami, Florida, USA, December 6-9, 2009, pages: 1094-1099.
  • (ICCS-09) G. Nie, G. Wang, P. Zhang, Y. Tian, Y. Shi. Finding the Hidden Pattern of Credit Card Holder's Churn: A Case of China. International Conference on Computational Science , Baton Rouge, Louisiana, USA, May 25-27, 2009, pages: 561-569.
  • (ICCS-09) Y. Wang, P. Zhang, G. Nie, Yong Shi: Multiple Criteria Quadratic Programming for Financial Distress Prediction of the Listed Manufacturing Companies. International Conference on Computational Science , Baton Rouge, Louisiana, USA, May 25-27, 2009, pages: 616-624.
  • (ICCS-09) Y. Zhang, P. Zhang, Y. Shi. Kernel Based Regularized Multiple Criteria Linear Programming Model. International Conference on Computational Science , Baton Rouge, Louisiana, USA, May 25-27, 2009, pages: 625-632.
  • (BIFE-09) L. Zhang, J. Li, A. Li, P. Zhang, G. Nie, and Yong Shi. A New Research Field: Intelligent Knowledge Management. The Second IEEE International Conference on Business Intelligence and Financial Engineering , Beijing, China, July 24-26, 2009, pages: 450-454.
  • (CSIE-2009) G. Wang, G. Nie, P. Zhang, and Y. Shi. Personal Financial Market Segmentation Based on Clustering Ensembles. 2009 World Congress on Computer Science and Information Engineering , Los Angeles, USA, March 31-April 2, 2009, pages: 694-698.
  • (BIFE-09) P. Zhang, L. Zhang, G. Nie, Y. Zhang, Y. Shi, Transfer Knowledge via Relational K-Means Method, The Second IEEE International Conference on Business Intelligence and Financial Engineering , Beijing, China, July 24-26, 2009, pages: 656-659.
  • (BIFE-09) X. Zhou, Y. Shi, P. Zhang, G. Nie, W. Jiang: A New Classification Method for PCA-Based Face Recognition. The Second IEEE International Conference on Business Intelligence and Financial Engineering , Beijing, China, July 24-26, 2009, pages: 445-449.
  • (MCDM-09) P. Zhang, Y. Tian, D. Zhang, X. Zhu, and Y. Shi, A Multiple Criteria and Multiple Constraints Mathematical Programming Model for Classification. International Conference on Multiple Criteria Decision Making , Chengdu, China, June 21-26, 2009, pages: 600-605.
  • (MCDM-09) P. Zhang, X. Zhu, G. Wang, L. Zhang, and Y. Shi, Mining Knowledge from Multiple Criteria Linear Programming Models, International Conference on Multiple Criteria Decision Making , Chengdu, China, June 21-26, 2009, pages: 170-175.
  • (MCDM-09) G. Wang , F. Li , P. Zhang, Y. Tian, and Yong Shi,Data Mining for Customer Segmentation in Personal Financial Market,International Conference on Multiple Criteria Decision Making ,Chengdu, China, June 21-26, 2009, pages: 614-621.
  • (WI-08) P. Zhang, Y. Tian, Z. Zhang, A. Li, and X. Zhu. Select Objective Functions for Multiple Criteria Programming Classification. Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology s, Sydney, Australia, December 9-12, 2008, pages: 420-423.
  • (WI-08) D.Zhang, Y. Tian, P. Zhang. Kernel-Based Nonparametric Regression Method. Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology s, Sydney, Australia, December 9-12, 2008, pages: 410-413.
  • (WI-08) M. Zhu, Y. Shi, A. Li and P. Zhang. A Bias-Variance Analysis of Multiple Criteria Linear Programming Classification Ensembles. Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology s, Sydney, Australia, 2008.
  • (WI-08) L.Zhang, J. Wei, A. Huang, J. Li, and P. Zhang. DEA-Based Comprehensive Evaluation of Intelligent Knowledge. Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology s, Sydney, Australia, December 9-12, 2008.
  • (ICCS-08) P. Zhang, Y. Tian, X. Li, Z. Zhang, and Y. Shi. Select Representative Samples for Regularized Multiple-Criteria Linear Programming Classification. International Conference on Computational Science , Krakow, Poland, June 23-25, 2008, pages: 436-440.
  • (ICCS-08) Z. Zhang, Y. Shi, P. Zhang, and G. Gao. A Rough Set-Based Multiple Criteria Linear Programming Approach for Classification. International Conference on Computational Science, Krakow, Poland, June 23-25, 2008, pages: 476-485.
  • (KAM-08) A. Li, J. Su, M. Zhu, P. Zhang, A Quantitative Study on the Mutual Fund Rating with LDA,The 2nd International Symposium on Knowledge Acquisition and Modeling , Wuhan, China, November 30-December1, 2009, pages: 606-609.
  • (ICDM-07) P. Zhang and J. Dai, Multiple Criteria Linear Programming for VIP E-Mail Behavior Analysis. Proceedings of the 7th IEEE International Conference on Data Mining, Omaha, Nebraska, USA, October 28-31, 2007, pages: 289-296.
  • (ICDM-07) Y. Shi, Y. Tian, X. Chen and P. Zhang. A Regularized Multiple Criteria Linear Program for Classification, Proceedings of the 7th IEEE International Conference on Data Mining, Omaha, Nebraska, USA, October 28-31, 2007, pages: 253-258.
  • (ICCS-07) P. Zhang, J. Zhang, and Y. Shi. A New Multi-criteria Quadratic-programming Linear Classification Model for VIP E-Mail Analysis, International Conference on Computational Science , Beijing, China, May 27-30, pages: 499-502.

    Research 4: Other Directions

    My research also spans over information security, data quality management and information retrieval.

  • (ISI-13) Y.Shang, P.Zhang, Y.Cao, L.Guo, Behavioral Targeting With Social Regularization. In Proceedings of the 2013 IEEE Intelligence and Security Informatics , June 4-7,2013, Seattle Washington, USA.
  • (Journal) B. Gao, D. Buttler, D. Anastasiu, S. Wang, P. Zhang, and J. Jan. User-centric Organization of Search Results. IEEE Internet Computing, 2012.
  • (ITQM-13) W. Zang, P. Zhang, X. Wang, J. Shi, and L. Guo, Detecting Sybil Nodes in Anonymous Communication Systems, the 2013 International Conference on Information Technology and Quantitative Management , China, 2013.
  • (NAS-13) Y.Shang, P.Zhang, Y.Cao, A New Interest-sensitive and Network-sensitive Method for User Recommendation. In Proceedings of the 8th IEEE International Conference on Networking, Architecture, and Storage , July 17-19, 2013, Xi'an, Shanxi, China.
  • (RAID-13) P. Fu, G. Xiong, Y. Zhao, M. Song and P. Zhang, An Identification Method Based on SSL Extension, Research in Attacks, Intrusions and DefensesSymposium, St. Lucia, October 23-25, 2013.
  • (GCC-10) J. Cai, Z. Zhang, P. Zhang, X. Song: Rethinking the Building Block: A Profiling Methodology for UDP Flows. GCC 2010: 332-337
  • (WI-10) J. Yan, X. Yun, P. Zhang, J. Tan, L. Guo: A New Weighted Ensemble Model for Detecting DoS Attack Streams. Web Intelligence/IAT s 2010: 227-230
  • (Journal) X. Li, L. Zhang, P. Zhang and Yong Shi. Problems and systematic Solutions in Data Quality, International Journal of Services Sciences. Vol. 2 (1), 2009, pages: 53 -69.
  • (Journal) P. Zhang, Z. Zhang, A. Li, and Y. Shi, Global and Local Bagging Approach for Classifying Noisy Dataset, International Journal of Software and Informatics, Vol.2(2), 2008, pages: 181-197.
  • (WI-08)L.Zhang, J. Wei, A. Huang, J. Li, and P. Zhang. DEA-Based Comprehensive Evaluation of Intelligent Knowledge. Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology s, Sydney, Australia, December 9-12, 2008.
  • (Journal) X. Li, Y. Shi, J. Li and P. Zhang. Data Mining Consulting Improve Data Quality, Data Science Journal, Vol. 6 (2007), pages: 658-666.

    System 1: SocialAnalysis: A Real-time Twitter Stream Query and Mining System

    We designed a real-time system on Twitter to discover and summarize emergent social events from twitter streams. Social events, such as parade and riots, are characterized with a common phenomenon that people always participate in each procedure, which contains preparations, organization and initiation for the event. As the development of social networks, people always frequently post messages or comments about their activities and opinions. Hence, there exist temporal correlations between the physical world and virtual social networks, which can help us to monitor and track social events, detecting and positioning anomaly events before their outbreakings, so as to provide early warning. The key technologies in the systems involve: (1) Data denoising methods based on multi-features, which screens out the related specific event data from massive data. (2) Abnormal events detection methods based on statistical learning, which can detect anomalies by analyzing and mining series of observation and statistics on the time axis. Because time series data have features of high velocity, high noise and nonlinearity, we designed models to calculate the anomaly scores for each time point in order to do fast detection. When the anomaly score at some certain time point is higher than a threshold, the point will be considered as an abnormal node. (3) Geographical position recognition, which is used to recognize regions where abnormal events may happen. Because social events may also occur in multiple, clustering methods such as density based method are used to obtain the approximate geographical locations, and then the maximum likelihood estimation can be used to accurately locate location by considering trusted weights.

Image

Visit the system demo : http://192.168.122.180:8080/WeiBo/ (Currently for IIE inner access only.)

 

    System 2: OPTMINER: A Convex Optimization Toolbox for Classification

    The OPTMINER toolbox was designed for data classification based on multiple criteria mathematical programming. It has been successful applied to credit card portfolio management, bioinformatics, fraud management, intrusion detection and email log data analysis. It realizes the auto-process of data classification by sealing the details of middle operations, including parameter adjustment and the access of intermediate results. Users only need to feed data with format complying with the original data file (database). Users can get the statistical results of the density distribution, KS Value and the parameters. In contrast to OPTMiner1.0, OPTMiner2.0 are more intelligent in recognizing the data source, it can automatically calculate the file / database table size and the number of attributes, so that users don't need to manually input these parameters. OPTMiner2.0 can automatically fill up missing attribute values. When the algorithm encounters an error pop-up warning window, users are suggested to modify the existing source data errors according to the warning information.

Image

For DOS program here, for Windows Program here. Windows requires Microsoft .NET Framework 2.0.

Locations of visitors to this page