Kiri Wagstaff: Research

Research

Full list of publications (chronological)

KEDS (Knowledge-Enhanced Discovery Systems): This project is a collaboration with Marie desJardins of UMBC. Generally, we are working to develop classification, clustering, and preference modeling methods that can take advantage of existing domain knowledge. Funded by a five-year grant from the NSF's Information Technology Research Program (NSF award number 0325329).
- Learning User Preferences for Sets of Objects (pdf, 8 pages, 270k).
  Marie desJardins, Eric Eaton, and Kiri L. Wagstaff.
  Proceedings of the Twenty-Third International Conference on Machine Learning, 2006.
- Active Constrained Clustering by Examining Spectral Eigenvectors (pdf, 14 pages, 681Kb).
  Qianjun Xu, Marie desJardins, and Kiri L. Wagstaff.
  Proceedings of the Eighth International Conference on Discovery Science, 2005.
- DD-PREF: A Language for Expressing Preferences Over Sets (pdf, 7 pages, 597Kb).
  Marie desJardins and Kiri L. Wagstaff.
  Proceedings of the Twentieth National Conference on Artificial Intelligence, p. 620-626, 2005.
- Constrained Spectral Clustering under a Local Proximity Structure (pdf, 2 pages, 55Kb).
  Qianjun Xu, Marie desJardins, and Kiri Wagstaff.
  Proceedings of the 18th International Florida Artificial Intelligence Research Society (FLAIRS) Conference, 2005.
- A Context-Sensitive and User-Centric Approach to Developing Personal Assistants (pdf, 3 pages, 40Kb).
  Marie desJardins, Eric Eaton, and Kiri Wagstaff.
  Persistent Assistants: Living and Working with AI (AAAI Spring Symposium), 2005.
- Making the Most of Missing Values: Object Clustering with Partial Data in Astronomy (pdf, 5 pages, 85Kb).
  Kiri Wagstaff and Victoria G. Laidler.
  Astronomical Data Analysis Software and Systems XIV (ADASS), ASP Conference Series, Vol. ???, 2005.
- Clustering with Missing Values: No Imputation Required (pdf, 10 pages, 244Kb).
  Kiri Wagstaff.
  Classification, Clustering, and Data Mining Applications (Proceedings of the Meeting of the International Federation of Classification Societies), p. 649-658, 2004.
HARVIST (Heterogeneous Agricultural Research Via Interactive, Scalable Technology): collaboration with Dominic Mazzoni of JPL and Stephan Sain of CU Denver. We are developing and demonstrating a machine learning analysis toolkit that uses SVMs, clustering, and multivariate spatial models to identify the connections between weather and agriculture (e.g., crop yield). We will integrate data from orbiting satellites, weather stations on the ground, land cover types, soil properties, and historical crop yield archives. Funded by a two-year grant from NASA's Earth Science Technology Office (grant number AIST-QRS-04-3004).
- Recent HARVIST Results: Classifying Crops from Remote Sensing Data (pdf, 5 pages, 197k).
  Kiri Wagstaff and Dominic Mazzoni.
  Proceedings of the Second NASA Data Mining Workshop, 2006.
- HARVIST: A System for Agricultural and Weather Studies Using Advanced Statistical Models (pdf, 5 pages, 3.5Mb).
  Kiri L. Wagstaff, Dominic Mazzoni, and Stephan Sain.
  Proceedings of the Earth-Sun Systems Technology Conference, 2005.
Classification of Martian Surface Units and Representative Regions of Interest: This project is a collaboration with Scott Murchie of APL. My portion of the project involves using constrained clustering to create a map of Mars surface materials. Funded by a two-year grant from NASA's Mars Data Analysis Program (grant number NNG04GJ076).
- Automated Analysis of Mars Multispectral Observations (pdf, 3 pages, 195Kb). (Abstract/poster)
  Kiri Wagstaff and James F. Bell III.
  Sixth International Mars Conference, 2003.
Automated Identification of Mars Polar Caps: This project is funded by the New Millenium Program and NASA's Intelligent Systems Program. We developed a method for autonomously identifying the seasonal CO2 polar caps on Mars. I have also done some work on automated detection of the mysterious Martian dark slope streaks.
- An Onboard Data Analysis Method to Track the Seasonal Polar Caps on Mars (pdf, 8 pages, 2.8Mb).
  Kiri L. Wagstaff, Rebecca Castano, Steve Chien, Anton B. Ivanov, and Timothy N. Titus.
  Proceedings of the International Symposium on Artificial Intelligence, Robotics, and Automation in Space, 2005.
- Towards Onboard Orbital Tracking of Seasonal Polar Volatiles on Mars (pdf, 2 pages, 115Kb). (Abstract/poster)
  Kiri L. Wagstaff, Rebecca Castano, Steve Chien, Anton B. Ivanov, and Timothy N. Titus.
  36th Lunar and Planetary Science Conference, 2005.
- Tracking Retreat of the North Seasonal Ice Cap on Mars: Results from the THEMIS Investigation (pdf, 2 pages, 860Kb). (Abstract/poster)
  Anton B. Ivanov, Kiri L. Wagstaff, and Timothy N. Titus.
  36th Lunar and Planetary Science Conference, 2005.
Hyperspectral Data Prioritization: This is an ongoing project in our group at JPL, funded by the Interplanetary Network Directorate. Last year, I worked on generating science priority maps for 2D images, as an input to a region-of-interest compression algorithm developed at JPL (ROI-ICER) and, in collaboration with Henry Shu (an undergraduate at Caltech who worked with me for the summer), extended my constrained clustering methods to incorporate spectral libraries when analyzing hyperspectral data.
This year, my focus is on priority maps for 3D compression (a much more challenging problem!) and methods for selecting the best subset of images for transmission, given one or more mission science goals.
- Semi-Supervised Data Summarization: Using Spectral Libraries to Improve Hyperspectral Clustering (pdf, 8 pages, 635k).
  Kiri L. Wagstaff, Henry P. Shu, Dominic Mazzoni, and Rebecca Castano.
  The Interplanetary Network Progress Report, vol. 42-163, 2005.
- Validating Rover Image Prioritizations (pdf, 8 pages, 354Kb).
  Rebecca Castano, Kiri Wagstaff, Lin Song, and Robert C. Anderson.
  The Interplanetary Network Progress Report, vol. 42-160, 2005.
- Current Results from a Rover Science Data Analysis System (pdf, 10 pages, 696Kb).
  Rebecca Castano, Michele Judd, Tara Estlin, Robert C. Anderson, Daniel Gaines, Andres Castano, Ben Bornstein, Tim Stough, and Kiri Wagstaff.
  Proceedings of the 2005 IEEE Aerospace Conference, 2005.
- Using Trained Pixel Classifiers to Select Images of Interest (pdf, 8 pages, 262Kb).
  Dominic Mazzoni, Kiri Wagstaff, and Rebecca Castano.
  The Interplanetary Network Progress Report, vol. 42-158, 2004.
- How to Leverage Existing Spectral Knowledge when Clustering Hyperspectral Data. (Abstract/poster)
  Kiri Wagstaff, Henry Shu, and Rebecca Castano.
  Fall AGU, 2004.
- Science-Based Region-of-Interest Image Compression (pdf, 2 pages, 68Kb). (Abstract/poster)
  Kiri Wagstaff, Rebecca Castano, Sam Dolinar, Matt Klimesh, and Ryan Mukai.
  35th Lunar and Planetary Science Conference, 2004.

Last modified: Wed Jun 14 00:26:03 2006