Conference on Intelligent Data Understanding 2012
Invited Speakers

Invited Speakers:

Methodology

Vasant Honavar



Program Director, Information Integration and Informatics at NSF



Professor, Iowa State University

Dr. Vasant Honavar received his Ph.D. in Computer Science and Cognitive Science in 1990 from the University of Wisconsin Madison, specializing in Artificial Intelligence. Since 1990, he has been on the faculty of Iowa State University (ISU) where he is currently a Professor of Computer Science and of Bioinformatics and Computational Biology. He directs the Artificial Intelligence Research Laboratory (which he founded in 1990) and the Center for Computational Intelligence, Learning & Discovery (which he founded in 2005). Honavar’s research has resulted in foundational contributions in machine learning algorithms and applications, bioinformatics and computational biology, knowledge representation, and data, knowledge, and process integration. He has published over 225 research articles in peer-reviewed journals and conferences. Honavar is a senior member of the Association for Computing Machinery (ACM), and of the Institute of Electrical and Electronic Engineers (IEEE) and a member of the Association for Advancement of Artificial Intelligence (AAAI), and a member of the Board of Directors for ACM Special Interest Group on Bioinformatics. Honavar is currently on an assignment as a Program Director for the Information Integration and Informatics Program within the Division of Information & Intelligent Systems of the Computer and Information Science and Engineering Directorate of the National Science Foundation where his programmatic include the Big Data Science and Engineering (BIGDATA), the Information Integration and Informatics, and the Smart Health and Wellbeing Programs. Honavar has received numerous awards and honors during his career including the several best paper awards, Director’s Award for Collaborative Integration for his work at NSF, the Margaret Ellen White Graduate Faculty Award, the Iowa Board of Regents’ Award for Faculty Excellence, and College of Liberal Arts and Sciences Award for Research Excellence at Iowa State University. However, his proudest accomplishments are the large number of PhD students that he has had the pleasure and good fortune to have worked with during his career.

Talk Title: Learning Predictive Models from Large Distributed Autonomous Data Sources

Recent advances in high throughput data acquisition, distributed sensors, and networked information systems offer unprecedented opportunities in collaborative, integrative data analysis (e.g., for discovery of a priori unknown complex relationships, construction of predictive models from data), hypothesis generation, and knowledge creation. However, realizing these opportunities presents several challenges in practice: Data and knowledge repositories are often autonomous, large, and distributed. Semantic differences, differences in scope, intended use, and privacy considerations further complicate their effective use in practice. In this talk, I will summarize some recent progress on algorithms for constructing predictive models from distributed, semantically disparate data in settings where centralized access to data is be neither feasible nor desirable. I will briefly outline some approaches to selective reuse of knowledge from multiple autonomous knowledge bases; and the automated composition of autonomous software services into complex workflows. I will conclude the talk with some open research challenges in Discovery Informatics that need to be addressed in order to be able to fully realize the promise offered exponential growth in the volume, velocity, and variety of data in scientific discovery. Much of this research has been carried out in collaboration with current and former members of the Iowa State University Artificial Intelligence Research Laboratory and has been supported in part by grants from the National Science Foundation.

Elizabeth Bradley



Professor, University of Colorado

Elizabeth Bradley did her undergraduate and graduate work at MIT, interrupted by a one-year leave of absence to row in the 1988 Olympic Games, and has been with the Department of Computer Science at the University of Colorado at Boulder since January of 1993. Her research interests include nonlinear dynamics, artificial intelligence, and
control theory. She is the recipient of a NSF National Young Investigator award, a Packard Fellowship, a Radcliffe Fellowship, and the 1999 student-voted University of Colorado College of Engineering teaching award.

Talk Title: Chaos, Computers, and Nonlinear Dynamics

Nonlinearity and chaos are ubiquitous and fascinating. Chaotic systems, in particular, are exquisitely sensitive to small perturbations, but their behavior has a fixed and highly characteristic pattern. Understanding this somewhat counterintuitive combination of effects is important to one's ability to model the physical world. I will begin this talk by reviewing of some of the basic ideas of the field of nonlinear dynamics and describe how those ideas can be leveraged to analyze time-series data. Most of these
nonlinear time-series analysis techniques were developed for
low-dimensional systems, however, and many of them require perfect models --- situations that are rare in the geosciences. For practitioners in these fields, then, it is important to understand how and when to use nonlinear time-series analysis, how to interpret the results, and how to recognize when and why these methods fail. I will
demonstrate all of this in the context of a specific problem: understanding and predicting processor and memory loads in modern multi-core computers.

Earth and Environmental Systems

Bill Mahoney



Deputy Directory, Research Applications Laboratory, NCAR

Mr. Mahoney is the Deputy Director of the Research Applications Laboratory (RAL) at the National Center for Atmospheric Research (NCAR), in Boulder, Colorado. He has been involved in research and development activities at NCAR for more than 25 years and has directed weather research and development programs in aviation, surface transportation weather, social sciences, agriculture, verification, intelligent forecast systems, and renewable energy.

Mr. Mahoney received his M.S. degree in Atmospheric Science from the University of Wyoming in 1983 and received his B.S. degree in Aeronautics from Miami University of Ohio in 1981. He spent more than a decade researching microburst wind shear and its impact on aviation operations and was involved in the development of the Low-Level Wind Shear Alert System (LLWAS) and the Terminal Doppler Weather Radar (TDWR). More recently his focus has been on developing advanced surface transportation weather hazard detection capabilities and wind energy prediction technologies.

In addition to his program management duties, Mr. Mahoney is involved in program development and commercialization activities at NCAR.

Mr. Mahoney has written or co-authored more than 40 papers and frequently presents NCAR’s work at national and international conferences and seminars. He is a member of the American Meteorological Society (AMS), Intelligent Transportation Society of America (ITSA), American Wind Energy Association (AWEA), and the Utility Variable Generation Integration Group (UVIG), and he is active on several atmospheric science and surface transportation committees.

Mr. Mahoney is the Past-Chair of the AMS Board on Enterprise Economic Development, was a member of the AMS Economic Development Committee, and is an AMS Fellow.

Talk Title: Big Earth – Big Data!

The electronic age has provided enormous opportunities to advance our understanding of the world around us. It has allowed us to work remotely but collaboratively, generate and move terabytes of data at levels never anticipated even two decades ago, bring science to those who never could have imagined having the opportunity to contribute solutions, and it has allowed us to explore the details and intricate workings of nearly every part of our physical world. But our ability to process all this information is outpacing our infrastructure and is requiring us to develop new analytical methods and techniques that can fully exploit these datasets. The geoscience community is dealing with these same issues and society is now demanding detailed historical, current and future information about our Earth system. This talk will discuss these challenges and opportunities.

Marika Holland



Climate and Global Dynamics Division, NCAR

Dr. Marika M. Holland is a Scientist III in the Oceanography Section of the Climate and Global Dynamics Division of NCAR’s Earth System Laboratory. Her research interests are focused on the role of sea ice in the climate system, including secular sea ice change, ice-ocean-atmosphere interactions, abrupt high latitude climate change, and polar climate variability. Dr. Holland currently serves as Chief Scientist for the Community Earth System Model (CESM) project and previously served as co-chair for the CESM Polar Climate Working Group. She has contributed to sea ice model developments for the Community Earth System Model, including work that enables the quantification of the role of black carbon deposition on Arctic sea ice loss. Dr. Holland has also been an active member of numerous committees and advisory panels for the Arctic Research Consortium of the U.S., the National Science Foundation, and the National Academy of Sciences among others. She has been a contributing author on the Intergovernmental Panel on Climate Change third and fourth assessment reports and contributed to numerous other national and international assessments on the changing Arctic climate.

Talk Title:
Investigation of the Climate System Using Earth System Models

Earth System models are complex numerical tools designed to study the Earth's climate system. Traditionally, climate models have focused on the physical climate system. More recently, these have transitioned to Earth System models, with the incorporation of new simulation capabilities. These enable new science on process interactions and feedbacks through the inclusion of active biogeochemical cycles, atmospheric chemistry and ice sheet components. In addition to increasing model complexity, there is an interest in better characterizing regional climate information on shorter prediction timescales. This has motivated a push towards higher spatial resolutions and initialized climate forecasts. The end result is that these modeling systems are producing increasingly large datasets for an increasing number of climate-science applications. I will discuss the models themselves, the typical simulation output that they produce, and some of the problems to which they are being applied. The challenges inherent in climate science and earth system modeling will be addressed. Finally, the new problems and opportunities that arise from an ever-increasing volume of simulation output will be discussed.

Lawrence Buja and Caspar Ammann



NCAR

Dr. Lawrence Buja is the Director of the Climate Science and Applications Program (CSAP) at the National Center for Atmospheric Research (NCAR) in Boulder, Colorado, which carries out interdisciplinary
research on social, economic, and political activities related to
climate at local, regional and global scales. CSAP addresses impacts,
adaptation and vulnerability to climate change by generating scenarios
of projected climate change, developing tools and methods for analyzing
current and future vulnerability, and conducting integrated analyses of
climate change impacts and adaptation. Dr. Buja also serves as the scientific project manager for the Climate
Change and Prediction group. This group carried out the climate
simulations of the earth's past, present and future climate with NCAR's
Community Climate System Model (CCSM) that made up the joint US NSF/DOE
submission to the Intergovernmental Panel on Climate Change (IPCC).
Together with the CCSM runs carried out on the Japanese Earth Simulator,
these CCSM runs were the largest data submission to the IPCC AR4 effort
by any modeling center in the world. In addition to carrying out IPCC
climate simulations, Lawrence is a contributing author to both the 2001
IPCC Third Assessment Report and the breakthrough 2007 IPCC Fourth
Assessment Report. Lawrence also works closely with the World Bank, the InterAmerican
Development Bank and other international agencies applying NCAR’s
climate and regional model expertise to help guide sustainable
development investment strategies throughout the developing world.

Talk Title: Big-Data Enabling Usable Climate Science and Services for Society at the Petascale

Massive peta-scale climate model and observation data
archives have been integrated by the international climate research
community into a global distributed virtual data center in support of
the ongoing IPCC Coupled Model Intercomparison Project research programs.
Sophisticated big-data
approaches will be required to efficiently bring this rich data resource
to bear on the new ”Climate 2.0” question of understanding and
quantifying the impact of climate change on the coupled human-natural
system. On one side, one can screen these data holdings through data
mining with increasingly smart algorithms that look for systematic or
emerging patterns and events. On the other side, rapid access to these
archives provides an opportunity for application-specific vulnerability
and risk assessments through direct and efficient selection and
propagation of carefully designed scenarios from the simulations and
models. NCAR/UCAR, in collaboration with the university science and
cyber-infrastructure communities, is the right place to establish such
Scenario Theaters to bring the climate science to applications and let
decision makers explore the robustness of their management instruments
in a realistic climate change context. This presentation will describe
current trends in global/regional climate science & data and how we hope
to work with big-data specialists to explore and integrate across
multiple complex physical and social science data streams to assess and
mitigate the impact of extreme weather and climate events on vulnerable
societal sectors.

Space Sciences

Fran Bagenal



Professor, University of Colorado

Dr Fran Bagenal was born and grew up in England. In 1976, inspired by NASA’s missions to Mars and the prospect of the Voyager mission, she came to the US for graduate study at MIT. Her 1981 PhD thesis involved analysis of data from the Voyager Plasma Science experiment in Jupiter’s giant magnetosphere. She spent 1982-1987 as a post-doctoral researcher in space physics at Imperial College, London. Voyager flybys of Uranus and Neptune brought her back to the US and she joined the faculty at the University of Colorado, Boulder in 1989. She is professor of Astrophysical and Planetary Sciences and faculty associate of the Laboratory of Atmospheric and Space Physics.

In addition to the Voyager mission, Dr Bagenal has been on the science teams of the Galileo mission to Jupiter and the Deep Space 1 mission to Comet Borrelly. She edited Jupiter: Planet, Satellites and Magnetosphere (Cambridge University Press, 2004). She heads the plasma teams on the first two New Frontiers missions: New Horizons mission to Pluto (launched January 2006) and Juno, a Jupiter polar orbiter.

Dr Bagenal has served on several committees of the National Research Council of the National Academy of Sciences: Space Studies Board, Committee on Planetary and Lunar Exploration, Solar and Space Physics Decadal Survey Committee, and chaired the Committee to Assess the Role of Solar and Space Physics in Exploration. She has also chaired NASA’s Planetary Science Subcommittee and the Outer Planets Assessment Group.

Dr Bagenal became a US citizen on 9/6/2001 and Fellow of the American Geophysical Union in 2006.

Talk Title: Juno: Mission to Jupiter's Interior - and Poles

The mystery of how much water resides in the interior of Jupiter is a major issue for understanding the formation of our solar system - as well as giant planets around other stars. The Juno mission aims to determine Jupiter's interior structure via magnetic and gravity sounding. Scanning in six bands of microwaves will ascertain the abundance and distribution of water. Juno’s orbit over Jupiter’s poles is designed to allow the spacecraft to map Jupiter’s gravity and magnetic fields and the amount of water in its atmosphere, but the polar vantage point also affords Juno a perfect opportunity to study this completely unexplored region of magnetosphere. Some of the charged particles in the magnetosphere are funneled into the polar atmosphere to create intense auroral emissions, which Juno will observe with unprecedented resolution. Instruments on the spacecraft will measure the flux particles that interact with the atmosphere to generate the auroras. Ultraviolet and infrared images will provide visual context for data from particles and fields instruments which will elucidate how charged particles are accelerated to 10s of keV energies in Jupiter's magnetosphere.The Juno spacecraft was launched in August 2011 and will reach Jupiter in 2016 where it will go into an eccentric polar orbit, skimming over the clouds and under the hazardous radiation belts. Returning science data from the outer solar system via the Deep Space Network will be at rates of ~15 Gb per orbit. The challenges are to spread these capabilities between 9 instruments and, once the data reaches the ground, turning them into maps of the interior or comparing them with magnetospheric simulations.

Greg Laughlin



Professor, University of California, Santa Cruz

Gregory Laughlin is Professor of Astronomy and Astrophysics at the University of California, Santa Cruz. He did his undergraduate work at the University of Illinois, and obtained his Ph.D. from UC Santa Cruz. In the mid-1990s, he was an NSF-JSPS Fellow at the National Observatory of Japan, and did postdoctoral work at Berkeley and the University of Michigan. Prior to joining the UCSC faculty in 2001, he was a Space Scientist at the NASA Ames Research Center in Mountain View California. Laughlin's primary interest is the detection and characterization of extrasolar planets, and his research encompasses both theoretical and data-intensive work. He has made use of the Spitzer Space Telescope to characterize time-dependent weather on short-period planets, and has done global hydrodynamic modeling to interpret the data. He is the co-designer of the Systemic Console software which is used by many researchers to model time series data and to perform N-body simulation. He also writes a web log, www.oklo.orghttp://www.oklo.org, that provides an overview of emerging issues in the field of extrasolar planet research.

Talk Title: Big Data from the Galactic Planetary Census

Fifteen years ago, the salient features of the known extrasolar planets could be written down on an index card. At present the catalog of extrasolar planets numbers in the thousands, and the rate of detection is increasing rapidly. Highly diverse planets are being identified through a diverse set of observational techniques; photometric transit detection, Doppler radial velocimetry, gravitational microlensing, and direct detection via adaptive optics imaging are all producing discoveries at an increasing rate. In this talk, I will give an overview of the census as currently understood, and I will show how the different detection methods are producing complementary detections. Like many areas in astronomy, exoplanetary detection is facing issues related to "Big Data". Large online repositories (such as that produced by the Kepler Mission) serve many terabytes of data, much of which has gone analyzed due to the time-consuming algorithms required. The talk will seek to highlight the current issues, and will show how ad-hoc collaborations across the community are being formed to deal with the challenges (and the excitement) of this fast-moving area.

Aerospace and Engineering Systems

James Kuchar



MIT Lincoln Laboratory

Dr. James K. Kuchar is Leader of the Weather Sensing Group at MIT Lincoln Laboratory, focusing on integrated sensing and decision-support systems for transportation safety and efficiency. Recently he has helped manage the development of prototype automation systems such as the Tower Flight Data Manager, Route Availability Planning Tool, and dynamic replanning capabilities for US Transportation Command’s logistics network. Dr. Kuchar previously worked on safety analysis and algorithm design for the Traffic Alert and Collision Avoidance System (TCAS) and modeling and simulation to support access of unmanned aircraft into civil airspace. Dr. Kuchar joined Lincoln Laboratory in 2003 after having been on the MIT faculty for eight years in the Department of Aeronautics and Astronautics, where he also received his SB, SM, and PhD degrees. A private pilot, he has authored over 60 journal and refereed conference papers.

Talk Title: Experiences from Modeling and Exploiting Data in Air Traffic Control

Recent machine learning techniques are now enabling significant advances in the performance of air transportation decision support systems. This talk will review three vignettes from recent data-driven prototype system development: exploiting radar data and modeling airspace traffic encounters to build a more effective collision avoidance system, extracting information from surface surveillance data to improve airport operations, and learning from operational experience to enhance the Route Availability Planning Tool and facilitate departure management in the vicinity of convective weather. In each case, examples and challenges of data collection, processing, and translation into models and ultimately operational prototype systems will be discussed.