Transcript Slide 1

FODAVA-Lead Education, Community Building, and
Research:
Dimension Reduction and Data Reduction:
Foundations for Interactive Visualization
Haesun Park and Guy Lebanon
School of Computational Science and Engineering
Georgia Institute of Technology
FODAVA Review Meeting, Dec. 9, 2010
FODAVA-Lead PIs at GAtech
Alex Gray
Associate Director
CSE
Machine Learning
Fast Algorithms for Massive DA
Industry Relations
Vladimir Koltchinskii
Mathematics
Machine Learning Theory
Computational Statistics
Haesun Park
Director
CSE, Associate Chair
Numerical Computing
Data Analysis
Research
FODAVA Community Building
John Stasko
Associate Director
IC, Associate Chair
Information Vis.
Collaboration with NVAC and DHS/CoE
Liaison with Vis. community
Renato Monteiro
ISyE
Continuous Optimization
Statistical Computing
FODAVA-Lead Senior Personnel
James Foley
Graphics and Visualization, HCI
Visual Analytics Digital Library
Alexander Shapiro
ISyE
Stochastic Programming
Optimization
Multivariate Stat. Analysis
Richard Fujimoto
Associate Director
CSE, Chair
Modeling and Simulation
Education and Outreach
Santosh Vempala
CS
Theory of Computig
Director of ARC
Guy Lebanon
Associate Director
CSE
Machine Learning
Computational Statistics
Education
Hongyuan Zha
CSE
Numerical Computing
Data Analysis
Director of Graduate Studies
Arkadi Nemirovski
ISyE
Optimization
Non-parametric Stat.
Hao-Min Zhou
Mathematics
Wavelet and PDE
Image Processing
FODAVA-Lead Mission
Research: Serve as a central facility to involve all FODAVA
awardees in a common effort to develop the scientific
foundations for data and visual analytics
Education: Facilitate the development of a body of
knowledge, curricula, and education programs to establish
and build DAVA workforce
Community Building:
•Integrate diverse DAVA communities and reach out for broader
participation
•Serve as a liaison between FODAVA researchers and NVAC, DHS
Centers of Excellence
FODAVA Curriculum Development and Education
•
•
Data and Visual Analytics Education Workshop:
Aug. 30, 2010, University of Maryland, College Park
Organizers: John Stasko, Haesun Park, Richard Fujimoto, Guy Lebanon (Georgia
Tech); David Ebert, Marti Burns, Tim Collins (Purdue); Georges Grinstein (U. Mass);
Richard May and Kris Cook (PNNL)
• Continued series of workshops e.g., VAST (Oct 2008), Georgia Tech (Dec 2008)
• Focus on experiences in visual analytics course and curriculum development
• Identified major topics in DAVA education programs
• Collaboration with NVAC, DHS Centers of Excellence
• Identified best practices and needs
• Sample course syllabi and curriculum
• Refined DAVA body of knowledge
• Resources, syllabi, and discussion available on vacommunity.org website
(http://www.vacommunity.org)
Development of a Course on Data and Visual Analytics on the interface
between data analysis and information visualization
•
•
Emphasis on practical methods and case studies
Development of Cluster of Core Graduate Courses in DAVA:
•
Data and Visual Analytics
•
Computational Data Analytics
•
Information Visualization
FODAVA Curriculum Development and Education
• Dissemination of homework assignments, exams, lecture notes,
demonstrations, etc. via an online blog on data and visual analytics
http://smlv.cc.gatech.edu/dava/
• Working with IEEE CIS Task Force on Data Visualization and Data
Analysis to disseminate visual analytics techniques and education to
the machine learning and data mining communities
http://www2.in.tu-clausthal.de/~hammer/DVA/
•
•
•
Workshop on 'Challenges of Data Visualization' at NIPS 2010
http://cseweb.ucsd.edu/~lvdmaaten/workshops/nips2010/
Special session on 'Perspectives of Dimensionality Reduction and Visual
Analytics' at CIDM 2011
Workshop on 'High-dimensional data visualization' at ICCS
FODAVA Outreach Program
• GT CRUISE Program (Computing Research
Undergraduate Intern Summer Experience)
 Encourage students to consider graduate studies
 Diverse student participation
- Multicultural, emphasizing minorities, women
- U.S. and international students
 Ten week summer research projects
 Interdisciplinary individual and group projects and CRUISE-wide
events
- Weekly seminars (technical, grad studies)
- Symposium: conference-style presentations
• Participation in VAST Challenge Competitions
•
•
VAST Challenge 2009 participation resulted in one award
VAST Challenge 2010 Participation resulted in two awards
DAVA Community Development
Visualization Community
• Extreme Scale Visual Analytics Workshop, IEEE VisWeek, Salt Lake City,
October 24, 2010 ( D. Ebert, G. Lebanon, P. McCormick, H. Park, H. Pfister,
and L. Wilkinson)
–
–
–
–
–
–
–
–
Invited talks by A. Gray, L. Wilkinson, W. Cleveland, M. Maggioni
Extremely high dimensional data
Real time, scalable computational methods
Programming support
Parallel and High Performance Computing
Speedup and accuracy tradeoffs
Fundamental limits and theory
Execution on computationally limited platforms
• Tutorial on Machine Learning for Information Visualization, IEEE VisWeek,
Salt Lake City, October 24, 2010 (G. Lebanon, F. Sha)
• Participation in NVAC Consortium Meetings, November 2008, August 2009,
August 2010
• Forum on Geometric Aspects of Machine Learning and Visual Analytics:
Recent Developments and Future Challenges, IEEE VisWeek, Atlantic City,
October 11-12, 2009 (M. Maggioni, V. Koltchinskii, A. Varshney, H. Park)
• Birds-of-Feather Session, VAST Conference, Columbus Ohio, October 2008
(K. Cook, K. Ma, and H. Park)
Additional IEEE VisWeek 2010 Activities
• Papers
– Z. Liu, J. Stasko, “Mental Models, Visual Reasoning and Interaction in Information Visualization: A
Top-down Perspect,” InfoVis, 2010.
– Z. Liu, J. Stasko, “The Role of Theory in Information Visualization" Theories in Information
Visualization: What, Why and How, InfoVis Workshop, 2010.
– J. Choo, H. Lee, J. Kihm, H. Park, “iVisClassifier: An Interactive Visual Analytics System for
Classification Based on Supervised Dimension Reduction,” IEEE VAST 2010.
• Posters
- R. Basole, M. Hu, P. Patel, J. Stasko, Visualizing Converging Business Ecosystems for
Competitive Intelligence, InfoVis 2010.
• Panel
- Challenges in Visualizing Biological Data: N. Gehlenborg, C. Gorg, M. Meyer, C. Nielsen, InfoVis 2010.
• VAST Challenge Competition: Two Awards
- Good Support for Data Ingest: Data Ingestion and Evidence Marshalling in Jigsaw (Z. Liu, C. Gorg, J.
Kihm, H. Lee, J. Cho, H. Park, J. Stasko)
- Excellent Process Explanation: GeneTracer: Gene Sequence Analysis of Disease Mutations
(H. Lee,
J. Choo, C. Gorg, J. Shim, J. Kihm, Z. Liu, H. Park, J. Stasko)
• Doctoral Colloquium: 2 students participated
• Conference Organization
- Infoviz Program Committee: Edward Clarkson, Carsten Gorg, John Stasko
- InfoViz Steering Committee, VizWeek Executive Committ: John Stasko
- IEEE VAST Program Committee: Haesun Park
DAVA Community Development
Data Analysis Community
• Statistical Machine Learning for Visual Analytics, NIPS Conference, Vancouver,
B.C., Canada, December 11, 2009 (G. Lebanon and F. Sha)
–
–
–
–
Text visualization, compression counting
Visualization using probabilistic models
Manifold learning, networks and hyper-graphs
Dimension reduction, visual analytics for audio
• Large-Scale Machine Learning: Parallelism and Massive Datasets, NIPS
Conference, Vancouver, B.C., Canada, December 11, 2009 (C. Guestrin, A. Gray,
A. Smola, A. Gretton, J. Gonzalez
–
–
–
–
–
–
–
Multicore / Cluster based Learning Techniques
Machine Learning on Alternative Hardware (GPUs, Cell Processor, FPGAs, iPhone, ...)
Distributed Learning
Learning results and techniques on Massive Datasets
Large Scale Kernel Methods
Fast Online Algorithms for Large Data Sets
Parallel Computing Tools and Libraries
FODAVA Kickoff and Annual Meetings, September 2008, December 2009,
December 2010
Additional Activities by FODAVA Team Members
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
IEEE VisWeek 2010 Poster Session: "Exploration & Representation of Data with Geometric Wavelets” (Eric E Monson, Rachael
Brady, Guangliang Chen & Mauro Maggioni)
VAC Consortium Meeting Poster/Demo Session ” (Eric E Monson, Rachael Brady, Guangliang Chen & Mauro Maggioni)
Invited lecture at IEEE VisWeek 2010 workshop (Mauro Maggioni)
A talk at the VAC Consortium 2009 (Mark Hasegawa-Johnson)
2010 Modern Massive Data Set Workshop at Stanford (Ping Li)
Structure Discovery in 3-D Point-Cloud Data. Industrial Light and Magic, San Francisco, September 2009. (Guibas)
Sensing Mobile Objects and Applications, NTT Laboratories, Kyoto, Japan, October 2009 (Guibas)
Structure Discovery in 3-D Geometry, SIAM/ACM Joint Conf. Geometric and Physical Modeling. San Francisco, October 2009.
Also, at KAUST, Jeddah, Saudi Arabia, March 2010. (Guibas)
Image Webs, A9 (Amazon) Research, Palo Alto, November 2009. (Guibas)
The Information is in the Maps, Google, Mountain View, December 2009. Also, at Microsoft Asia Research Center, Beijing,
China, May 2010. Also at Qualcomm Research, Santa Clara, July 2010. Also, at INRIA, Sophia-Antipolis, France, August 2010.
(Guibas)
Information Dissemination and Cross-Correlation under Mobility, Army Research Labs, Adelphi, MD, April 2010. (Guibas)
Voronoi Diagrams in Geometry Processing and Network Routing, International Symposium on Voronoi Diagrams in Science and
Engineering, Quebec City, Canada, June 2010. (Guibas)
The Structure of Isometric Maps and Symmetries, EPFL Bernoulli Symposium, Lausanne, Switzerland, August 2010. (Guibas)
Invited session organized by J. Li in Joint Statistical Meetings (JSM), Vancouver, Canada, July 2010: Statistical Modeling and
Learning for Information Visualization and Dimension Reduction (Jia Li)
“Mode Based Clustering with Applications to Information Visualization,” J. Li, X. Zhang, Penn State University (Jia Li)
Panel of Visualization and Rich Data Sets in the Annual Workshop of Human-Computer Interaction Consortium, February, 2010
(Jia Li)
Panel Presentation: Interactive Visualization of Large Data Sets: Challenges and Some Preliminary Answers (Jia Li)
An invited talk at NIPS 2010 workshop on Challenges of Data Visualization (Getoor)
An invited talk at NIPS 2010 workshop on Networks Across Disciplines: Theory and Applications (Getoor)
Invited talks at International Conference on Data Analysis (May 2010), NEH/IPAM Institute on Networks for the Humanities
(Aug 2010), and organized 2010 KDD Workshop on Mining and Learning from Graphs (Getoor)
Organized NSF III PI Meeting (April 2010) (Singh and Getoor)
Gave tutorials on Exploting Statistical and Relational Information on the Web and in Social Media: Applications, Techniques,
and New Frontiers AAAI10 and will be giving tutorials at WSM 2011 and SDM 2011 (Getoor)
Birds-of-a-Feather session at VisWeek 2010: Scalable Interactive Visualizations for Visual Analytics (Ted Selker and Ole J.
Mengshoel)
FODAVA Distinguished Lecture Series
• Lecture
series
featuring
leaders
in
the
DAVA
community
• Develop in collaboration with
FODAVA partners and NVAC
• Web-cast
2010 FODAVA DLS:
• William Ribarsky, ”Developing a Visual Analytics Approach to Analytic Problem
Solving” February 26, 2010
• Leeland Wilkensen, “The Mathematical Foundation of Analytical Visualization” April
2, 2010
• Jim Thomas, “Three I’s of Visual Analytics for FODAVA Teams: Interdisciplinary,
International, Immediacy” April 16, 2010
2009 FODAVA DLS:
• Alan Turner, William Cleveland, Joseph Kielman, Alexey Chervonenkis, Vladimir Vapnik
FODAVA Website
http://fodava.gatech.edu
• Information on FODAVA Projects
• Dissemination of FODAVA results to communities:
• FODAVA Tech Report Series
• Repository of Data Sets
• FODAVA meetings/lecture/DLS materials
• DAVA community events and meeting information
• Blog on DAVA Taxonomy and course material