Knowledge Discovery from Huge Data Ensemble by an Integration of Automatic Data Selection and Pattern Extraction

Japanese version

Grants-in-Aid for Scientific Research, Fundamental Research (B), 25280085

Japan Society for the Promotion of Science (JSPS)

2013/4/1 - 2016/3/31

Research organization

Einoshin Suzuki

Principal investigator, Professor

Department of Informatics, ISEE, Kyushu University

Research abstract: To discover knowledge from a large number of huge data sets, we invented, developed, and implemented 4 highly novel methods that integrate automatic data selection and pattern discovery. The method that discovers cluster distribution meta-patterns exhibited high recalls and precisions under difficult conditions of high noise contamination and ambiguous and mutually overlapping cluster boundaries and was proved to be time-efficient. The method that discovers directional non-zero weight meta-patterns, which is based on multi-task classification based on sparse modeling, showed its practicability on various kinds of data including facial expression data measured with Kinect. The method that hierarchically clusters linear classifiers and the method that evaluates and discovers general classification rules each holding true in its respective data set showed their effectiveness on various synthetic and real data.


Selected outcomes:
1. Daisuke Ikeda and Einoshin Suzuki: "Finding Peculiar Compositions of Two Frequent Strings with Background Texts", Knowledge and Information Systems, An International Journal, Vol. 38, No. 3, pp. 567-597, Springer, March, 2014. DOI 10.1007/s10115-012-0601-y
2. Shin Ando, Theerasak Thanomphongphan, Daisuke Hoshino, Yoichi Seki, and Einoshin Suzuki: "Ensemble Anomaly Detection from Multi-resolution Trajectory Features", Data Mining and Knowledge Discovery, Vol. 29, No. 1, pp. 39-83, Springer, January 2015. DOI 10.1007/s10618-013-0334-x
3. Einoshin Suzuki, Yutaka Deguchi, Daisuke Takayama, Shigeru Takano, Vasile-Marian Scuturici, and Jean-Marc Petit: "Towards Facilitating the Development of a Monitoring System with Low-Cost Autonomous Mobile Robots", Information Search, Integration and Personalization, pp. 57-70, Communications in Computer and Information Sciences, Vol. 421, Springer, 2014. DOI 10.1007/978-3-319-08732-0_5
4. Yutaka Deguchi, Daisuke Takayama, Shigeru Takano, Vasile-Marian Scuturici, Jean-Marc Petit, Einoshin Suzuki: "Multiple-Robot Monitoring System Based on a Service-Oriented DBMS", Proc. Seventh ACM International Conference on Pervasive Technologies Related to Assistive Environments (PETRA 2014), Rhodes Island, Greece, May 2014. DOI 10.1145/2674396.2674418
5. Yutaka Deguchi, Einoshin Suzuki: "Skeleton Clustering by Autonomous Mobile Robots for Subtle Fall Risk Discovery", Foundations of Intelligent Systems, Lecture Notes in Computer Science 8502 (ISMIS 2014), Springer-Verlag, pp. 500-505, June 2014, Roskilde, Denmark. DOI 10.1007/978-3-319-08326-1_51
6. Angdy Erna, Linli Yu, Kaikai Zhao, Wei Chen, and Einoshin Suzuki: "Facial Expression Data Constructed with Kinect and their Clustering Stability", Active Media Technology, Lecture Notes in Computer Science 8610 (AMT 2014), Springer-Verlag, pp. 421-431, August 2014, Warsaw. DOI 10.1007/978-3-319-09912-5_35
7. 田之上伸吾,鈴木英之進:生涯学習の人物表情分類問題における実験的評価,平成26年度(第67回)電気・情報関係学会九州支部連合大会,pp. 372-373,鹿児島,2014年9月.
8. Daisuke Takayama, Yutaka Deguchi, Shigeru Takano, Vasile-Marian Scuturici, Jean-Marc Petit, Einoshin Suzuki: "Multi-view Onboard Clustering of Skeleton Data for Fall Risk Discovery", Ambient Intelligence (AmI 2014), LNCS 8850, Springer-Verlag, pp. 258-273, Eindhoven, November 2014. DOI 10.1007/978-3-319-14112-1_21
9. Ryosuke Kondo, Yutaka Deguchi, Einoshin Suzuki: "Developing a Face Monitoring Robot for a Deskworker", Ambient Intelligence (AmI 2014), LNCS 8850, Springer-Verlag, pp. 226-241, Eindhoven, November 2014. DOI 10.1007/978-3-319-14112-1_19
10. Shin Ando, Einoshin Suzuki: "Discriminative Learning on Exemplary Patterns of Sequential Numerical Data", Proc. 2014 IEEE International Conference on Data Mining (ICDM 2014), pp. 1-10 Shenzhen, China, December 2014. DOI 10.1109/ICDM.2014.122
11. Shin Ando, Einoshin Suzuki: "Minimizing Response Time in Time Series Classification", Knowledge and Information Systems, An International Journal, Springer (accepted for publication) DOI 10.1007/s10115-015-0826-7.
12. Einoshin Suzuki: "Multi-Task Data Mining toward Automating the KDD Process", Sixth International Conference on Information Technology and Electrical Engineering (ICITEE 2014) + Regional Conference on Computer and Information Engineering 2014 (RCCIE 2014), Jogjakarta, October 2014 (keynote lecture).
13. Einoshin Suzuki, Yutaka Deguchi, Tetsu Matsukawa, Shin Ando, Hiroaki Ogata, Masanori Sugimoto: "Toward a Platform for Collecting, Mining, and Utilizing Behavior Data for Detecting Students with Depression Risks", Proc. Eighth International Conference on Pervasive Technologies Related to Assistive Environments (PETRA 2015), Corfu, Greece, July 2015. DOI 10.1145/2769493.2769538
14. Somar Boubou, A. H. Abdul Hafez, Einoshin Suzuki: "Visual Impression Localization of Autonomous Robots", Proc. 2015 IEEE International Conference on Automation Science and Engineering (CASE 2015), pp. 328-334, Gothenburg, Sweden, August 2015. DOI 10.1007/s10844-014-0329-0
15. Vasile-Marian Scuturici, Yann Gripay, Jean-Marc Petit, Yutaka Deguchi, Einoshin Suzuki: "Continuous Query Processing over Data, Streams and Services: Application to Robotics", New Trends in Databases and Information Systems (ADBIS 2015), pp. 36-43, Communications in Computer and Information Sciences (CCIS), Vol. 539, Springer, Poitiers, France, September 2015. DOI 10.1007/978-3-319-23201-0_5
16. Einoshin Suzuki: "On the Feasibility of Discovering Meta-Patterns from a Data Ensemble", Discovery Science (DS 2015), LNAI 9356, Springer-Verlag, pp. 266-274, Banff, Canada, October 2015. DOI 10.1007/978-3-319-24282-8_22
17. Kaikai Zhao, Einoshin Suzuki: "Clustering Classifiers Learnt from Local Datasets Based on Cosine Similarity", Foundations of Intelligent Systems, LNCS 9384 (ISMIS 2015), Springer-Verlag, pp. 150-159, Lyon, France, October 2015. DOI 10.1007/978-3-319-25252-0_16
18. Yutaka Deguchi, Einoshin Suzuki: "Hidden Fatigue Detection for a Desk Worker Using Clustering of Successive Tasks", Ambient Intelligence (AmI 2015), LNCS 9425, Springer-Verlag, pp. 263-283, Athens, November 2015. DOI 10.1007/978-3-319-26005-1_18
19. Yutaka Deguchi, Daisuke Takayama, Shigeru Takano, Vasile-Marian Scuturici, Jean-Marc Petit, Einoshin Suzuki: "Skeleton Clustering by Multi-Robot Monitoring for Fall Risk Discovery", Journal of Intelligent Information Systems, Vol. ?, No. ?, pp. ?-?, Springer (accepted for publication) DOI 10.1007/s10844-015-0392-1
20. Einoshin Suzuki: Compression-Based Evaluation of a Meta-Pattern in Terms of a Belief and a Data Ensemble, Abstracts and Timetable of the Twelfth International Conference on Operations Research (ICOR 2016), p. 50, Havana, Cuba, March 2016.