Sponsored by | 
|
The Computer Science Colloquium Series is a forum for invited speakers, faculty, and graduate students to share research ideas. Everyone is invited to attend and participate. An up-to-date listing of the speakers and abstracts of their talks will be posted here. Please e-mail the colloquium committee if you are interested in speaking or know of someone who would be a good addition to our program. Thank you.
Colloquia are generally held every Thursday at 3:40 p.m. except during academic holidays. See below for specific times and topics. Refreshments will be served after every colloquium in the conference room, 225 Atanasoff Hall.
In some cases, the colloquium will start at 4.10 pm and refreshments will be served earlier starting at 3.30 pm. These colloquiums are marked with an asterisk (*) below.
Listing of Talks
Several speakers have agreed to present but have not yet been scheduled. Potential dates for these talks are listed as "to be announced" in the table below. All other dates are open. Please contact one of us listed below if you are interested in speaking or know of a potential contributor to our series.
(Top)
Abstracts
1. MULTIAGENT SYSTEMS ENGINEERINGScott DeLoach
This presentation describes the work that Dr. DeLoach has been pursuing in the area of the specification, design, and development of multiagent systems. In multiagent systems, he is interested in coordinating the behavior of individual autonomous agents to provide a specific system-level behavior. The Multiagent Systems Engineering (MaSE) methodology developed by Dr. DeLoach uses the abstraction provided by multiagent systems for developing intelligent, distributed software systems. To accomplish this goal, MaSE uses a number of graphically based models to describe the types of agents in a system and their interfaces to other agents, as well as an architecture-independent detailed definition of the internal agent design. The primary focus of MaSE is to guide a designer from an initial set of requirements through the analysis, design, and implementation of a multiagent system that satisfies the original specification. This methodology is the foundation of the agentTool development system, which also serves as a validation platform and a proo for MaSE. The agentTool system implements all the models used in MaSE and is currently being extended to provide semi-automated transformation of system analysis artifacts into a design that is guaranteed to fulfill the initial system specification. Unlike many other multiagent methodologies, MaSE is independent of any particular agent architecture, programming language, or communication framework-the focus is on building heterogeneous multiagent systems. 2. High Performance Computing: building and running of inexpensive clusters capable of supercomputing performanceRicky Kendall
& colleagues
Nearly all scientific and engineering disciplines have forefront research requiring high performance computing for advanced simulation and modeling. An increasingly viable solution for small groups to acquire the resources needed for demanding computations is the building of clusters of workstations or PCs. Clusters are not only a capacity computational resource but they are ideal development systems for large scale parallel systems. While the need for parallelized algorithms and code may be a barrier, it is also one that needs to be overcome for utilization of modern supercomputers having thousands of processors. The Scalable Computing Laboratory (SCL) of the Ames Laboratory has been building and optimizing such clusters for many years as part of their research commitments. In partnership with the IPRT Center for Physical and Computational Mathematics (CPCM), research scientists and students have been working with many ISU groups from many disciplines in code parallelization and in the building of clusters. Partnerships with several vendors have enhanced this process.
We particularly encourage interested groups to send graduate students to attend this meeting if they are interested in applications demanding large amounts of computer cycles or the computer science issues around High Performance Computing.
We will describe support mechanisms available for helping groups get over the barrier and into parallel computing and provide some examples of recent successes.
Agenda: The need for high performance computing 10 min. ISU clusters and their performance 15 min. Applications 20 min. Where do we go from here 5 min. Informal discussions with interested parties 30 min.
3. "Learning with Costs"Dragos D. Margineantu
Cost-sensitive learning addresses the task of learning to make predictions and decisions when different types of errors have different associated costs (penalties). Existing learning algorithms were designed to treat all errors the same and to minimize the *number* of errors that are made. In most practical applications, however, different errors have different costs, and methods are required that can minimize the total cost of all errors. For example, in medical diagnosis, the cost of diagnosing a patient as healthy when he or she has a life-threatening disease is much higher than the cost of making the opposite mistake.The talk will first outline the issues that need to be considered in designing and evaluating learning methods for the minimization of total cost. Then I will review the current approaches to constructing learning algorithms in a cost-sensitive context. Finally, I will introduce new methods for building and evaluating cost-sensitive learning systems and present experiments that compare different cost-sensitive methods on both real-world and synthetic data sets.
4. Separation of NP-completeness notionsPavan Aduri
The theory of reductions is one of the foundational pillars of complexity theory. The crucial notion of NP-completeness depends on reductions. Reductions translate instances of one problem to instances of another problem. Researchers have identified different notions of polynomial-time reductions, for example Turing or adaptive reductions, many-one reductions, and truth-table or non-adaptive reductions to study relations among different problems. Each reduction defines a completeness notion in NP. If these completeness notions are different in NP then P is not equal to NP follows easily. So it is natural to ask whether we can compare the relative strengths of these completeness notions under a reasonable hypothesis. For many years, this question remained open. Recently, Lutz and Mayordomo using the hypothesis "p-measure of NP is not zero" and Ambos-Spies and Bentzein under the hypothesis "NP has a p-generic language" proved that many of these completeness notions are mutually different. However the question about whether Turing completeness is different from truth-table completeness in NP remained open.In this talk we introduce a new hypothesis, different from measure and genericity hypotheses, and prove from this hypothesis that Turing completeness is different from truth-table completeness in NP. Using a weaker hypothesis we prove that Turing and many-one completeness notions differ in NP.
This is a joint work with Alan Selman.
5. Broadcast Routing with Minimum Wavelength Conversion in WDM Optical NetworksLu Ruan
As the traffic on Internet increases exponentially during the past decade, Wavelength Division Multiplexing (WDM) optical networks with terabits per second bandwidth become a natural choice for the next generation Internet backbone. In WDM optical networks, wavelength converters are placed at the routing nodes to reduce the blocking probability of connections. Since wavelength conversion needs O-E-O conversion that causes long delay, it is desirable to minimize wavelength conversions in supporting a connection in the network. Some network applications require the establishment of broadcast connections. The examples are weather reports distribution, stock market updates and live radio programs, etc. We study the problem of supporting a broadcast connection with minimum wavelength conversions using a graph model. We proved that the problem is NP-hard and therefore heuristic approach is needed to find near-optimal solutions. We designed a greedy approximation algorithm to solve the problem and showed that it achieves nearly-the-best performance ratio. We also conducted experiments to evaluate the algorithm using randomly generated network topologies. The results showed that the algorithm produces good approximation solutions to the problem. 6. Fast, Low-Cost Checkpointing and Recovery Techniques for Mobile Computing SystemsMukesh Singhal
Checkpointing and failure recovery techniques that have low overhead and provide fast recovery from failures are integral to the design of fault-tolerant, high-performance mobile computing systems. This talk will present a new approach called the quasi-synchronous checkpointing and failure recovery for mobile computing systems. The checkpointing algorithm preserves process autonomy by allowing them to take checkpoints asynchronously and uses communication-induced checkpointing for progression of the recovery line which helps bound rollback propagation during a recovery. Thus, it has easeness and low overhead of asynchronous checkpointing and recovery time advantages of synchronous checkpointing. There is no extra message overhead involved during checkpointing and the additional checkpointing overhead is nominal. The algorithm ensures the existence of a recovery line consistent with the latest checkpoint of any process at all time. The recovery algorithm exploits this feature to restore the system to a state consistent with the latest checkpoint of a failed process. The recovery algorithm has no domino effect and a failed process only needs to rollback to its latest checkpoint and request other processes to roll back to a consistent checkpoint. To avoid domino effect altogether, selective pessimistic message logging at the receiver end is used. 7. * SABRE: Software Architecture Based Requirements Engineering
Carl Chang
Major software errors or failures may be caused by the poor performance of requirements engineering, which is closely related to system engineering with overlapping concerns and common techniques.On the other hand, software architecture is another critical concern in the front-end software engineering process. Software architectural design requires a substantial amount of domain knowledge and often engages artifact reuse. The fact is that requirements engineering and software architecture are closely interrelated and complementary to each other. It has been known that it is naive to separate the two by treating requirements engineering only as the problem-space issue and software architecture only as the one belonging to the solution-space. Unfortunately, many authors and speakers often failed to treat the two in an integrated manner, resulting in philosophies and techniques that may not be fully applicable to the software industry.
For those who are aware of the existence of such a relationship, there generally lacks a complete methodology to bridge the process gap between the two. Software Architecture Based Requirements Engineering, abbreviated as SABRE, is the technology with emphasis on the software development and management processes where requirements engineering and software architectural design can be performed in a systematic and cohesive manner. The talk is about SABRE with both functional and object-oriented analysis and design principles in mind. A Function-Class Decomposition (FCD) method, an essential concept to SABRE, will be presented.
8. APT AgentsDiana F. Gordon
Intelligent agents, such as robots and softbots, are becoming increasingly prevalent. Furthermore, there is a movement toward having multiple agents cooperate in order to achieve complex tasks. The usefulness of such agents increases dramatically if they are able to adapt to unforeseen circumstances. Nevertheless, the price of added adaptability is typically a decrease in predictability. Our research addresses the important topic of behavioral assurance for adaptive multi-agent systems. We present the novel APT agents paradigm, which provides a foundation for developing agents that are adaptive to unforeseen circumstances, predictable in the sense of obeying critical constraints, and timely in their responses. In this paradigm, it is assumed that agents are engaged in a cooperative task. They adapt by applying learning operators to their plans. To ensure the preservation of critical constraints, including global constraints on multi-agent interactions, agents' plans are initially verified and repaired as needed. Furthermore, they are re-verified after every adaptation. The primary problem addressed by this research is the time efficiency of reverification. We present positive results that certain learning operators are a priori guaranteed to preserve useful classes of constraints, as well as very efficient reverification algorithms for those learning operators that have negative a priori results. Potential applications include adaptive but reliable communication and power networks, assurance for co-evolved multi-robot behaviors such as shepherding or tracking and docking, and coordinated behavior of planetary exploration rovers. 9. Physicomimetics for Controlling Swarms of AgentsWilliam M. Spears
Currently, a plethora of approaches are being developed to address the issue of collective/emergent behavior for large numbers of robots or other mobile agents. Unfortunately, current approaches are either practical (but ad hoc, i.e., not based on scientific principles), or theoretical (but difficult to apply). To address this, we propose a principled but practical approach to the control of large collections of agents, that we call physicomimetics. The essence of physicomimetics is that agents communicate with each other via virtual physics forces that are inspired by real physical forces. Using these virtual forces, agents can perform a variety of tasks, such as surveillance, perimeter defense, or configuring themselves into a distributed sensor grid. Because the forces are physics-based, agent collectives share the advantages of real particle aggregations, e.g., self-assembly, robustness, and self-repair. Furthermore, this enables us to draw upon the enormous body of literature in physics, chemistry, crystallography and mathematics to not only design our systems (synthesis), but to provide theoretical guarantees about emergent behavior (analysis). Such guarantees can make swarms of agents acceptable for real-world applications, where behavioral assurances are critical. 10. * Aspects of analysis and visualization of multidimensional datasetsRaj Acharya
This talk is comprised of three parts:(a) We will describe a software system for the identification of structures in images. We will demonstrate the use of the software system to identify anatomical structures in human radiographs. The system employs an object oriented knowledge model to perform the desired analysis.
(b) The software system will be used to analyze the genomic structure and function in multidimensions. We will present results, which support a dynamic mosaic model for the higher order arrangement of genomic function inside the cell nucleus.
(c) We will present a methodology for temporal analysis of complimentary DNA microarray gene expression data sets. 11. Detecting Group Differences
Stephen Bay
I will discuss the problem of detecting and describing the differences between several contrasting groups from observational data. These groups can be different classes of objects, such as male or female students, or the same group over time, e.g., freshman students in 1993 through 1998.My goal is to find contrast sets: conjunctions of attributes and values that differ meaningfully in their probability distribution across the groups. The main research issues are dealing with the enormous search space, controlling the search error to limit false discoveries, and summarizing the results for an end user. Throughout the talk, I will present examples based on an analysis of student record data from UC Irvine. I will also discuss some applications for a tool that finds contrast sets in characterizing the errors a classifier makes, multivariate discretization, and spatial clustering.
12. * Leader Election in Oriented Star NetworksPradip Srimani
In this talk we propose a leader election algorithm for the oriented star graph, where each edge in the topology is labeled with its dimension, but the nodes do not use any cannonical labeling. The algorithm exchanges O(N)messages and uses O(n^2log n) time where N=n! is the number of nodes in the star graph S_n. Time complexity of the algorithm is less than O(log N)^2 which is better than the corresponding known algorithm for hypercubes. 13. * Generic Approaches to Content-Based Retrieval of Visual InformationBill Grosky
The management of visual information is becoming quite important. Many companies have visually-based assets which need to be indexed, stored in a database, and queried. We have developed some generic and powerful approaches to represent visual objects. Our techniques are computational geometry based and are used to represent a visual object's point feature map, which is the spatial arrangement of an object's point-based features. In this talk, we discuss how we have used these approaches for shape-based and color-based features of image objects. 14. WWWPal - A System for Analysis and Synthesis of Web PagesM.S.Krishnamoorthy
WWWPal is a system that helps in the analysis and synthesis of Web documents. The system eliminates a common problem of obscure organization of Web documents in Web information systems. WWWPal consists of the following six components: 1) Web Robot, 2) Graph Visualizer, 3) Graph Analyzer, 4) Clustering Tool, 5) Synthesizer, and 6) Interface Package. The Web Robot is used to visit all the URLs that can be accessed from a given URL in a given Web server and to construct a graph. Graph Visualizer is used to display graphs (up to 200,000 nodes) in a variety of ways and also to edit a graph. Graph Analyzer is used to report different problems such as unreachable nodes, HTML documents that contain formatting errors, and broken links. The Clustering Tool is used to partition the nodes of the graph in different clusters using several heuristics. The partitioned graph is used to get better overall site maps of the Web server. In addition, this Clustering Tool helps in the visualization of large graphs. The Synthesizer helps to create a skeletal HTML document from a given graph. The Synthesizer interfaces with ASHE (A Simple HTML Editor) to edit a document. An interface package enables our system, WWWPal, to communicate with a browser, such as Netscape. This interface and Graph Visualizer form a skeletal graph browser (of the URLs) in our system.In this talk, I will also talk about two XML applications, XGMML (graph description language) and LOGML (reports description language) and a data mining application.
15. Tools for Parallelizing Serial Fortran Codes for Distributed Memory and Shared Memory Parallel ComputersCos Lerotheou
Dr. Cos Lerotheou, Professor ("Reader") of High Performance Computing from the High Performance Computing Group at the University of Greenwich, England will give a lecture about CAPTools. CAPTools is an interactive toolkit for the semi-automatic parallelization of serial Fortran codes developed by the Parallel Processing Group at the University of Greenwich. For more information see http://captools.gre.ac.uk/A few years ago, NASA-Ames did a world-wide evaluation of tools for helping researchers parallelize their codes. They found that the work being done at the University of Greenwich to be the most useful work being done in this area. NASA-Ames is currently collaborating with this group.
Sponsored by the Office of the Academic Information Technologies, the Department of Mathematics, and the Department of Computer Science.
16. HOPS: A Distributed Hybrid Optimization Technique for Protein Structure PredictionAlberto Maria Segre
The key to understanding the mechanism of life lies in understanding how proteins work. Nearly all functional aspects of an organism rely on proteins; enzymes, brain chemicals like dopamine, hormones, and hundreds of thousands of others. Surprisingly, a properly working protein works because it has just the right three dimensional shape, a shape determined by the protein's molecular composition, which is in turn described in the genome. Given that we now have access to extensive genomic information, the next challenge for computational biologists is to determine a protein's three dimensional shape (or ``tertiary structure'') -- and, consequently, its biological function -- from its molecular composition (or ``primary structure''), expressed as the sequence of constituent amino acids. This ''protein folding problem'' is enormously difficult, both because of the number of possible configurations a protein might assume and because we don't yet precisely understand the science of the folding process itself.We have been working on a new hybrid optimization approach to this problem that marks the convergence of several different research efforts. Our approach blends a distributed AI search technique we originally developed for use in automated deduction systems with a number of continuous optimization methods and powerful biochemically-inspired heuristics based on experimental data obtained in the laboratory. In this talk, I will describe the general architecture of our system, give an update on our recent progress, and demonstrate some preliminary folding results.
Joint work with Yinyu Ye (Management Sciences/Applied Mathematics), Kenneth Murphy (Biochemistry), Mauro Leoncini (CNR, Pisa, Italy), and Giovanni Resta (CNR, Pisa, Italy)
17. Dynamic Data Structures for Parallel PrologDesh Ranjan
Supporting efficient execution of Prolog and other declarative programming languages entails design and implementation of efficient dynamic data structures. In this talk, I will provide an overview of the data structure design problems that arise in this context and provide some of our solutions. I will also outline some related fundamental problems and our results for these problems. Although, the data structures we design are motivated by the needs of parallel logic programming they are very general and are of interest for general algorithm design.The talk will be self-contained.
18. ITR: Information Technology Revolution, Information Technology ResearchGeorge Strawn
NSF support for computer science research has dramatically increased under the auspices of a recent Presidential Initiative called ITR (Information Technology Research). ITR is interesting both for its proposed agenda of scientific research and for the process that resulted in the Administration proposing and Congress then appropriating funds for it. ITR has just begun the second year of a projected five-year ramp-up in Federal IT research support. To complete this ramp-up, it is now necessary that the new Administration make it a part of the President's budget request to Congress. This talk will: give a brief history of the ITR Initiative,highlight its scientific goals, indicate its successes (and challenges) thus far, and conclude with a prognosis for its continued successSpeaker Biographies
Scott DeLoachDr. DeLoach is currently an Assistant Professor of Computer Science and Engineering at the Air Force Institute of Technology (AFIT), where he has been since 1998. His research interests include specification, design and synthesis of multiagent systems, knowledge-based software engineering, and formal specification acquisition. At part of his research program at AFIT, he has advised 12 MS students and has served as a member of 3 PhD committees; written over 20 research related papers; and has brought in over $300,000 in research grants from the Air Force Office of Scientific Research, the Dayton Area Graduate Institute, and the Air Force Research Laboratory. Prior to coming to AFIT, Dr. DeLoach was the Technical Director of Information Fusion Technology at the Air Force Research Laboratory from 1996 to 1998. From 1987 to 1993 he was Chief of Systems Engineering and Electronic Combat Support at Headquarters Strategic Air Command and was a Computer Resources Engineer at Wright-PattersoEngineering from Iowa State University in 1982 and his MS and PhD in Computer Engineering from the Air Force Institute of Technology in 1987 and 1996.
Visit Scott DeLoach's hompage here.
Ricky Kendall& colleagues
Ricky Kendall(Moderator)
Dave Halstead,Dave Turner,Brett Bode, Mark Gordon,Bruce Harmon
The Scalable Computing Laboratory was created in 1989 as a joint effort of the Department of Energy through Ames Laboratory and Iowa State University through the Center for Physical and Computational Mathematics.
The mission of the Scalable Computing Laboratory (SCL) is to improve parallel computing through clustering techniques for use in scientific and engineering computation. Our goal is the delivery of supercomputing power at a fraction of the cost of traditional systems.
For more information and an overview of the SCL please see: http://www.scl.ameslab.gov
Visit Ricky Kendall
& colleagues's hompage here.
Name: Dragos D. Margineantu
Born on Aug.19, 1970 in Timisoara, Romania
Parents: High school teachers
(father: mathematics; mother: literature)
Education: "Diploma de Licenta" (5-year degree, M.S. equivalent,
1995) in Computer Science and Engineering from the Technical University of
Timisoara, Romania (currently named "Politehnica" University of
Timisoara)
Thesis: "3D Modeling and Volume Determination of Prostate from Ecographic
Images" (advisor: Prof. Stefan Holban)
Ph.D. in Computer Science (expected, April 2001) from Oregon State University
Thesis: "Methods for Cost-Sensitive Learning" (advisor: Prof. Thomas
G. Dietterich)
Research interests: Artificial intelligence, machine learning, adaptive systems, data mining. Ensemble learning, cost-sensitive learning. Approximation techniques in learning and artificial intelligence. Applied machine learning.
Honors, Affiliation, Achievments:
- Member of the Program Committee of the Eighteenth International Conference on
Machine Learning (ICML-2001)
- Co-organizer of the Workshop on "Cost-Sensitive Learning" in
conjunction with the Seventeenth International Conference on Machine Learning
(ICML-2000) - Stanford University, July 2000.
- Member of the Upsilon Pi Epsilon honor society in computer science
- reviewer for several artificial intelligence and machine learning journals
- recipient of several prizes at national high school level mathematics contests
in Romania
Personal interests: photography, classical music.
Visit Dragos D. Margineantu's hompage here.
Pavan AduriPavan Aduri is currently a Ph.D student at University at Buffalo. He is working under the guidance of Dr. Alan Selman. His areas of interest include complexity theory and algorithms. He obtained his Master's from IIT, Kanpur and Bachelors from JNTU, Hyderabad.
Visit Pavan Aduri's hompage here.
Lu Ruan Lu Ruan received the B.E. degree in computer science from Tsinghua University, Beijing, China, in 1996. She received the M.S. degree in computer science from the University of Minnesota-Twin Cities in 1999. She is currently a Ph.D. candidate at the Department of Computer Science and Engineering, University of Minnesota-Twin Cities and will reveive her Ph.D. in May 2001. Her research interests include computer networks and analysis and design of optimization algorithms.Visit Lu Ruan's hompage here.
Mukesh SinghalMukesh Singhal is a Full Professor of Computer and Information Science at The Ohio State University, Columbus. He received a Bachelor of Engineering degree in Electronics and Communication Engineering with high distinction from University of Roorkee, Roorkee, India, in 1980 and a Ph.D. degree in Computer Science from University of Maryland, College Park, in May 1986. His current research interests include wireless networks and mobile computing, computer networks, operating systems, database systems, distributed systems, performance modeling, and computer security. He has published over 140 refereed articles in these areas. He has coauthored two books titled ``Advanced Concepts in Operating Systems", McGraw-Hill, New York, 1994 and ``Readings in Distributed Computing Systems", IEEE Computer Society Press, 1993. He is a Fellow of IEEE. He is currently serving in the editorial board of "IEEE Trans. on Knowledge and Data Engineering" and "Computer Networks". He is also serving as a book series editor for a book series on Distributed Computing Systems for the Oxford University Press. He served as the Program Chair of the 6th International Conf. on Computer Communications and Networks, 1997 and of the 17th IEEE Symposium on Reliable Distributed Systems, 1998. He is currently serving as the Program Director of Operating Systems and Compilers program at National Science Foundation.
Visit Mukesh Singhal's hompage here.
Carl ChangChang is currently Director of the International Center for Software Engineering, University of Illinois at Chicago. He served as the Editor-in-Chief for IEEE Software from 1991-94. He is the co-founder of the IEEE International Conference on Requirements Engineering (ICRE) and chaired its steering committee from 1998-2000. He has published numerous technical papers in areas of requirements engineering, specification and verification, testing, software management, object-oriented metrics, net-centric computing, and distributed systems.
As a constant speaker to the industry, he consulted major industrial companies including: AT&T, Lucent Technologies (Bell Labs), Northrop Grumman, Motorola, Fujitsu, NEC, Institute for Information Industry (i.e. III-Taiwan), Corelis Technologie (France), etc., and conducted joint research with some of these companies. Chang received his PhD degree in computer science from Northwestern University. Prior to his current teaching and research job, he worked at GTE Automatic Electric and Bell Labs. Chang has been a volunteer leader to IEEE Computer Society for almost 20 years. He is a Fellow of IEEE and currently serves on the Executive Committee of IEEE Computer Society. He was elected to be the Secretary of Board of Governors in 1998 and appointed the Vice President for Press Activities in 1999. Most recently he was elected First Vice President and appointed to chair the Educational Activities Board in 2001. He is influential (as co-chair from 1998-2000) to the Computing Curricula 2001 project (www.computer.org/education/cc2001), a joint task force between IEEE-CS and ACM, also sponsored by NSF.
Visit Carl Chang's hompage here.
Diana F. GordonDr. Diana Gordon is currently a research scientist at the Naval Research Laboratory Artificial Intelligence Center. She received a Ph.D. in computer science from the University of Maryland in 1990 and has over 40 publications. She is known internationally for her research in machine learning, and her primary focus has been on adaptive agents. Recently, she pioneered a novel field of research -- formal verification of adaptive agents' plans. Dr. Gordon is a member of AAAI, Sigma Xi and the American Mathematical Society. William M. Spears
Dr. William Spears received his Ph.D. in computer science from George Mason University in 1998 and has been a research scientist at the Naval Research Laboratory Artificial Intelligence Center since 1985. Dr. Spears has over 40 publications, an international reputation as an expert in evolutionary algorithms, and has just published a book on this topic. His current interests, which include the principles of self-organization, self-repair and emergent properties in complex systems, led to the origination of physicomimetics. Dr. Spears is a member of Sigma Xi and the American Mathematical Society. Raj Acharya
Raj Acharya received his PhD degree from the University of Minnesota/Mayo Graduate School of Medicine in 1984. He is currently Professor and Chair of Computer Science & Engineering Department at SUNY Buffalo. Prior to joining SUNY Buffalo, he was a Research Scientist at Thomson (GE) Paris, France.
Dr. Acharya's research interests are in the general areas of Experimental Computer Science with applications in Multimedia Computing, Bioinformatics, Computational Biology and Fractals. His research work has been featured, among others in Businessweek, The Scientist, Mathematics Calendar, Diagnostic Radiology and the Biomedical Engineering Newsletter.
Visit Raj Acharya's hompage here.
Stephen BayStephen D. Bay received his B.A.Sc. and M.A.Sc. from the Department of Systems Design Engineering at the University of Waterloo. He is currently completing his P.D. in the Department of Information and Computer Science at the University of California, Irvine. His research interests include data mining and machine learning.
Visit Stephen Bay's hompage here.
Pradip SrimaniPradip K. Srimani is Professor and Chair of Computer Science at Clemson University. He has previously served the faculty of India Statistical Institute, Calcutta, Gesselschaft fuer Mathematik und Datenverarbeitung, Bonn, West Germany, Indian Institute of Management, Calcutta, India and Southern Illinois University, Carbondale, Illinois, Colorado State University, Ft. Collins, Colorado, and Technical University of Compiegne, France. He is currently the Editor-in-Chief of IEEE Computer Society Press and is an associate Editor of IEEE Transaction on Data and Knowledge Engineering and a contributing editor of IEEE Software. His research interests include mobile computing, distributed computing, parallel algorithms, networks and graph theory applications. He is a co-editor of two books on software reliability and distributed mutual exclusion algorithms by IEEE CS Press. He has guest-edited special issues for IEEE Computer, IEEE Software, VLSI Design, Journal of Systems \& Software, and Journal of Computer \& Software Engineering, IEEE Transactions on Software Engineering, Parallel Computing, International Journal of Systems Science. He is a member of the ACM/IEEECS Steering Committee on Curricula 2001. He is a Fellow of IEEE and a member of ACM. Complete vita is available at www.cs.clemson.edu/~srimani/vita.pdf
Visit Pradip Srimani's hompage here.
Bill GroskyWilliam I. Grosky is currently professor and chair of the Computer Science Department at Wayne State University in Detroit, Michigan. Before joining Wayne State in 1976, he was an assistant professor of Information and Computer Science at Georgia Institute of Technology in Atlanta. His current research interests are in multimedia information systems, hypermedia, databases, and web technology. Grosky received his B.S. in mathematics from MIT in 1965, his M.S. in Applied Mathematics from Brown University in 1968, and his Ph.D. from Yale University in 1971. He has given many short courses in the area of database management for local industries and has been invited to lecture on multimedia information systems world-wide. Serving also on many database and multimedia conference program committees, he is currently the Editor-in-Chief of IEEE Multimedia, and on the editorial boards of the Journal of Database Management and Pattern Recognition.
Visit Bill Grosky's hompage here.
M.S.KrishnamoorthyM. S. Krishnamoorthy received his Ph.D from the Indian Institute of Technology, Kanpur, and is an associate professor of computer science at Rensselaer Polytechnic in Troy, New York. Since the late 80s, he has been working in the area of syntax directed document analysis and on table recognition. In the mid 90s, John Punin (his doctoral student) and he developed a WYSIWYG editor (ASHE) for HTML on UNIX platforms. Krishnamoorthy and his students have also developed share-ware Graph Visualization programs. He was one of the designers for the WWWPal system, a software system to analyze Web documents. The WWWPal system won the the best poster award at the WWW8 conference in 1999.
Visit M.S.Krishnamoorthy's hompage here.
Cos LerotheouBiography currently unavailable.
Alberto Maria SegreBiography currently unavailable.
Visit Alberto Maria Segre's hompage here.
Desh RanjanDesh Ranjan got a B.Tech degree in computer science from Indian Institute of Technology, Kanpur in 1987 and Masters' and Ph.D. in computer science from Cornell University in 1990 and 1992 respectively under the guidance of Prof. Juris Hartmanis. He spent one year as a a post-doctoral fellow at the Max Planck Institute for Computer Science, Saarbrucken in 1992-93 working in the Efficient Algorithms group led by Prof. Kurt Mehlhorn. Since then he has been at New Mexico State University where he is currently an Associate Professor in the Computer Science Department. Desh primary research interests are efficient algorithm design and computational complexity. He is currenly working on algorithm and data structure design and development in primarily two domains -- efficient parallel declarative programming language implementations and computational biology.
Visit Desh Ranjan's hompage here.
George StrawnBiography currently unavailable.
Visit George Strawn's hompage here.
Archives
(Top)
Contacts
Thank you for visiting this page. Please send your suggestions and comments to one of us in the Computer Science colloquium committee.
- Soma Chaudhuri . chaudhur@cs.iastate.edu
- Giora Slutzki. slutzki@cs.iastate.edu
- Mayuresh Vartak. mayuresh@cs.iastate.edu
(Top)
mayuresh@cs.iastate.edu