CS Colloquium Series @ UCY | (sponsored by |
 |
CS Colloquium Series - Department of Computer Science - University of Cyprus
The Department of Computer Science at the University of Cyprus holds
research colloquiums and social hours approximately once weekly. All
university students, faculty, and staff are invited to attend.
Notifications about new and upcoming events are automatically
disseminated to a variety of institutional lists. If you don't receive
these notifications, but want to get informed about upcoming colloquium
announcements, you can do the following:
|
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Christos Η. Papadimitriou |
Abstract (In Greek):
Η υπολογιστική έρευνα αλλάζει τις επιστήμες (τις φυσικές, μαθηματικές, και κοινωνικές επιστήμες, καθώς και τις επιστήμες ζωής) όχι μόνο με το να τις ενδυναμώνει αναλυτικά, αλλά κυρίως με το να τους παρέχει νέες οπτικές γωνίες που συχνά οδηγούν σε απρόβλεπτες διαισθήσεις. Τα παραδείγματα είναι πολλά: Οι κβαντικοί υπολογισμοί προσφέρουν το κατάλληλο πλαίσιο για την αμφισβήτηση και την επανεξέταση μερικών από τις πιο βασικές αρχές της κβαντικής φυσικής, ενώ η στατιστική μηχανική έχει βρει μια ισχυρή αναλογία για τις αλλαγές φάσεων στην αποδοτικότητα των πιθανοτικών αλγορίθμων. Στα μαθηματικά, το ερώτημα "P vs. NP" ανήκει πλέον στη λίστα των πλέον σημαντικών άλυτων προβλημάτων, ενώ στα οικονομικά, η μελέτη της υπολογιστικής πολυπλοκότητας οδηγεί σε αναθεώρηση των προβλέψεων για οικονομικές συμπεριφορές και επηρεάζει το σχεδιασμό οικονομικών μηχανισμών, όπως των δημοπρασιών. Τέλος, στη βιολογία, κάποια από τα πλέον θεμελιώδη προβλήματα, όπως η κατανόηση του εγκεφάλου και της εξέλιξης, μπορούν να επαναδιατυπωθούν παραγωγικά με υπολογιστικούς όρους.
Please notice that the talk will be given in Greek
Short Bio (In Greek):
Ο Χρίστος Παπαδημητρίου σπούδασε στο Εθνικό Μετσόβιο Πολυτεχνείο όπου έλαβε Δίπλωμα Ηλεκτρολόγου-Μηχανολόγου Μηχανικού το 1972. Στη συνέχεια ακολούθησε μεταπτυχιακές σπουδές στο Πανεπιστήμιο Princeton, όπου απέκτησε Διδακτορικό Δίπλωμα το 1976.
Ο Χρίστος Παπαδημητρίου είναι από το 1996 μέχρι σήμερα ο Καθηγητής της Έδρας C. Lester Hogan στο Τμήμα Ηλεκτρολόγων Μηχανικών και Επιστήμης Υπολογιστών, University of California at Berkeley. Διετέλεσε μέλος Ακαδημαϊκού Προσωπικού στα Πανεπιστήμια Harvard (Gordon McKay Επίκουρος Καθηγητής Επιστήμης Υπολογιστών, 1976-1978), Berkeley (Miller Fellow for Science, 1978-1979), M.I.T. (Επίκουρος Καθηγητής 1979-1981, και Αναπληρωτής Καθηγητής μέχρι το 1983), Stanford (Καθηγητής Επιστήμης Υπολογιστών και Επιχειρησιακής Έρευνας 1983-1988), και University of California at San Diego (Καθηγητής της Έδρας Irwin Mark and Joan Klein Jacobs, Τμήμα Επιστήμης και Μηχανικής Υπολογιστών, 1988-1995). Διετέλεσε Καθηγητής στο Εθνικό Μετσόβιο Πολυτεχνείο (Τμήμα Ηλεκτρολόγων Μηχανικών) κατά την περίοδο 1981-1988.
Το ερευνητικό έργο του Χρίστου Παπαδημητρίου καλύπτει ένα ευρύ φάσμα θεωρητικών προβλημάτων σε πολλές περιοχές της Επιστήμης Υπολογιστών. Η έρευνά του οδήγησε σε θεμελιώδη αποτελέσματα στη Θεωρία Αλγορίθμων, Θεωρία Πολυπλοκότητας, Συνδυαστική Βελτιστοποίηση και Θεωρεία Βάσεων Δεδομένων. Το ερευνητικό του έργο επεκτείνεται και σε άλλους επιστημονικούς κλάδους όπως Μαθηματική Οικονομική, Επιχειρησιακή Έρευνα, Θεωρία Παιγνίων, και αυτό έχει αποτελέσει τη βάση για ποικίλες πρακτικές εφαρμογές (τελευταία μάλιστα στο Διαδίκτυο).
Το ερευνητικό έργο του Χρίστου Παπαδημητρίου αποτελείται από περισσότερες των τριακοσίων δημοσιεύσεων σε έγκριτα περιοδικά και πρακτικά συνεδρίων. Έχει συγγράψει πέντε διδακτικά βιβλία τα οποία χρησιμοποιούνται ευρέως σε προπτυχιακό και μεταπτυχιακό επίπεδο. Έχει επίσης διατελέσει μέλος στις εκδοτικές επιτροπές πολύ σημαντικών περιοδικών της Επιστήμης Υπολογιστών όπως Journal of the ACM, Journal of Computer and Systems Sciences, Algorithmica, Theoretical Computer Science, Information and Computation, SIAM Journal on Computing, Journal of AI Research, SIAM Journal on Discrete Mathematics, και άλλα.
Ο Χρίστος Παπαδημητρίου έχει τιμηθεί με πολλές διακρίσεις για τη συνεισφορά του στην Επιστήμη Υπολογιστών. Είναι εταίρος της Αμερικανικής Ακαδημίας Τεχνών και Επιστημών (Η.Π.Α.) από το 2001. Είναι μέλος της Εθνικής Ακαδημίας Μηχανικής (Η.Π.Α.) από το 2002. Ο Χρίστος Παπαδημητρίου είναι ACM Εταίρος από το 2001 ως αναγνώριση της εξαιρετικής συνεισφοράς του στα πεδία της Θεωρίας Πολυπλοκότητας, Θεωρίας Βάσεων Δεδομένων και Συνδυαστικής Βελτιστοποίησης.
Το 2002, ο Χρίστος Παπαδημητρίου έλαβε το Donald E. Knuth Prize, το οποίο απονέμεται από τους οργανισμούς ACM και IEEE ως αναγνώριση εξαιρετικής συνεισφοράς στις θεμελιώσεις της Επιστήμης Υπολογιστών. Ειδικότερα, το βραβείο απονεμήθηκε στον Χρίστο Παπαδημητρίου ως αναγνώριση των επί πολλά έτη και κεφαλαιώδους σημασίας συνεισφορών του στις θεμελιώσεις της Επιστήμης Υπολογιστών. Ο Χρίστος Παπαδημητρίου συγκαταλέγεται στον κατάλογο Highly Cited (http://www.isihighlycited.com/) για την Επιστήμη Υπολογιστών με βάση το μεγάλο αριθμό αναφορών στο δημοσιευμένο έργο του. Τέλος, είναι Επίτιμος Διδάκτωρ των Πανεπιστημίων ETH Ζυρίχης (Sc. D. honoris causea), Μακεδονίας (Ph. D., honoris causea), Αθηνών (Ph.D., honoris causea) και Κύπρου (Ph.D., honoris causea).
 | Web: http://www.cs.ucy.ac.cy/colloquium/ |
 | Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
 | RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
Sponsor: The CS Colloquium Series is supported by a generous donation from  |
The Department of Computer Science and the Department of Electrical and Computer Engineering at the University of Cyprus cordially invite you to the colloquium entitled:
 |
Speaker: Dr. Shubu Mukherjee |
Abstract:
Cavium Networks is a provider of highly integrated semiconductor processors that enable intelligent networking, communications, storage, video and security applications. Cavium Networks offers a broad portfolio of integrated, software compatible processors ranging in performance from 10+ Mbps to 40Gbps that enable secure, intelligent functionality in Enterprise, Data-Center, Broadband/Consumer and Access and Service Provider equipment. This talk will describe some of the design choices around Cavium's multicore processors to illustrate how simpler cores can make a bigger difference in this market than massively out-of-order power-hungry processors.
Short Bio:
Dr. Mukherjee is widely recognized as one of the experts on architecture design for soft errors. He has made pioneering contributions towards the design of Redundant Multithreading (RMT) techniques, architectural vulnerability modeling for soft errors, creation of performance modeling infrastructure called Asim (jointly with Dr. Joel Emer), design of the Alpha 21364 interconnection network, and the creation of the first shared memory prediction scheme. In 2009, Shubu won the Maurice-Wilkes award for outstanding contributions to computer architecture. This is the highest award given to a mid-career architect. Prior winners include Dirk Meyer (CEO AMD), Bill Dally (Chief Scientist Nvidia), Steve Scott (CTO Cray), and Anant Agarwal (Prof MIT and Founder of Several Companies). Shubu is also a Fellow of IEEE and Distinguished Member of ACM. He was the General Chair of 2004 ASPLOS and will be the Program Chair for 2011 HPCA conferences. He wrote the seminal book on "Architecture Design for Soft Errors," which has been highly acclaimed by Microprocessor Report as well as researchers and practitioners. Shubu holds 25 patents and has 23 patents pending. He has written over 50 technical papers in top architecture conferences and journals. Currently, Dr Mukherjee is a Distinguished Engineer at Cavium Networks involved in architecting Cavium's next network processor. He is also Adjunct Faculty with the Indian Institute of Technology, Kanpur. In the past, Shubu Mukherjee was a Principal Engineer and Director of Intel's SPEARS Group (Simulation and Pathfinding of Efficient and Reliable Systems). The SPEARS Group was responsible for spearheading architectural change and innovation in the delivery of enterprise processors and chipsets by building and supporting simulation and analytical models of performance, power, and reliability. Shubu has taken 5 innovations in large-scale system monitoring, soft error tolerant micro-architectures, performance simulation, parallel simulation, and on-chip interconnect design from conception to implementation. These innovations have resulted in 100s of millions of dollars in increased revenue for Intel and Compaq, reduced internal costs by 10s of millions of dollars, influenced over a dozen products, and improved customer goodwill significantly.
| Additional Lecture by Dr. Mukherjee are listed here |
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus and ACM Cyprus cordially invite you to the colloquium entitled:
 |
Speaker: Prof. Yannis Ioannidis |
Abstract:
Human actions in real life are often influenced by several characteristics of the individual human involved in the actions. These characteristics can be broadly classified into three categories: those that are unique to the individual, those of the social environment of the individual, and those of the overall context or situation in which the individual is found while performing the actions. Usability of various types of information systems, e.g., database systems, digital libraries, or the Web, increases dramatically if the information they provide and their overall behavior is customized to these characteristics. Such personalization, socialization, and contextualization of information provision touches upon a broad spectrum of technical and other challenges. This talk describes the general problem and its associated challenges, hints upon a general framework for modeling a large number of cases, and offers some examples of systems and techniques that have been developed by the Univ. of Athens to address related challenges in various application environments.
Short Bio:
Yannis Ioannidis is currently a Professor at the Department of Informatics and Telecommunications of the University of Athens. He received his Diploma in Electrical Engineering from the National Technical University of Athens in 1982, his MSc degree in Applied Mathematics from Harvard University in 1983, and his Ph.D. degree in Computer Science from the University of California at Berkeley in 1986, and was on the faculty of the Computer Sciences Department of the University of Wisconsin at Madison, where he became a Professor before leaving in 1999. His research interests include database and information systems, electronic infrastructures, digital libraries, personalization, scientific systems, and human-computer interaction, topics on which he has published over one hundred articles in leading journals and conferences and holds three patents. Dr. Ioannidis is an ACM and IEEE Fellow and has received the Presidential Young Investigator (PYI) award in 1991, the 2003 VLDB "10-Year Best Paper Award", the 2006 nation-wide "Xanthopoulos-Pnevmatikos Award for Outstanding Academic Teaching" in Greece, and of several other teaching awards. He has been a program (co-)chair of ICDE'09 and several other conferences and a (co-)principal investigator in over thirty research projects funded by various government agencies (Europe, Greece, USA) or private industry. Dr. Ioannidis currently serves a 4-year term as the ACM SIGMOD Chair (following a 4-year term as Vice-Chair) and is or has been a member of several other executive bodies of professional organizations (VLDB Endowment, IEEE TCDE Executive Committee, EDBT Endowment) and Scientific Advisory Boards (Max Planck Institute for Informatics, Greek National Science & Technology Council, Information Technology advisor to the Greek Minister of Health).
|
Additional Lecture by Prof. Ioannidis:
|
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Stavros Harizopoulos |
Abstract:
Rising energy costs in large data centers drive the agenda for energy efficient computing. Towards this goal, it is critical to understand the interplay between energy consumption and performance in database servers. In the first part of this talk, I will focus on quantifying the role of database software in the overall energy efficiency of a server. Then, I will present the results of a recent study (SIGMOD'10) on the power usage profiles of database operators and I will explore the effect of different configuration parameters in the energy efficiency of a database system. Finally, I will discuss our work on query processing on solid state drives (SIGMOD'09), which have emerged as a primary building block for energy efficient storage.
Short Bio:
Stavros is an HP Labs researcher in the Intelligent Information Management Lab which is focused on enabling near real-time business intelligence with robust, scalable data management and data-intensive analytics. He received his Ph.D. in Computer Science from Carnegie Mellon, in 2005, and, through 2007, he worked as a Post-Doctoral researcher at the DB group of MIT. Stavros's research interests are in energy-efficient data management systems, query processing on new processor and storage technologies, main-memory transaction processing, and column-oriented databases.
For more information: http://nms.csail.mit.edu/~stavros/
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Dr. Didier Stricker |
Abstract:
It is has long been understood that automated systems aiming to assist or interact with human activity need to have a degree of understanding of human behaviour in order to be effective. Actions and responses need to align with our expectations and information needs to be presented in a manner which reflects our own perceptions. What is less well understood is how that understanding of behaviour is to be obtained.
In this talk we will present the very first work of the department "augmented vision" of DFKI. The focus lies on capturing technologies and includes head-, hand-, arm- tracking, object identification and scene reconstruction. The goal is to build a precise digital representation of a real and dynamic scene, including humans executing given tasks and interacting with the surrounding. The technologies involve visual-inertial sensor units, inertial on-body sensors, fisheye as well as high-resolution spherical HDR-images. Current results will be presented and discussed in relations with achieved quality and required computing resources.
Short Bio:
Didier Stricker is member of the Management Board of the German Research Center for Artificial Intelligence (DFKI GmbH) in Kaiserslautern where he leads the new research department "Augmented Vision". He is appointed as Professor at the Computer Science Department of the University of Kaiserslautern. From 2002 to June 2008 Didier Stricker lead the department "Virtual and Augmented Reality" at the Fraunhofer Institute for Computer Graphics (Fraunhofer IGD) in Darmstadt, Germany.
More information: http://av.dfki.de/
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Christos Papadopoulos |
Abstract:
Many Internet topology studies have appeared in the literature.
However, such studies have for the most part, ignored the population of
hosts. While many hosts are hidden behind firewalls and NATs, there is
much to be learned from examining the population of "visible" Internet
hosts -- one can better understand network growth and accessibility to
help assess vulnerabilities, deployment of new technologies and improve
network models.
This paper is to our knowledge the first attempt to measure the
population of visible Internet edge hosts. We measure hosts in two ways:
via periodic Internet censuses, where we query all accessible Internet
addresses every few months, and via surveys of a small fraction of the
responsive address space, probing each address every 11 minutes for
one week. These approaches are complementary: a census is effective at
evaluating the Internet as a whole, while surveys validate the census
and allow observation of the lifetime of typical address occupancy.
Our findings include trends in address occupancy, an upper bound on the
number of servers and an analysis of firewalled addresses and firewall
block size.
Joint work with John Heidemann, Yuri Pryadkin, Ramesh Govindan and
Joseph Bannister.
Short Bio:
Christos Papadopoulos is currently an associate professor at Colorado
State University. He received his Ph.D. in Computer Science in 1999
from Washington University in St. Louis, MO. His interests include
network security, router services, multimedia protocols and reliable
multicast. His current work includes signal processing techniques for
network attack detection and participation in the PREDICT program to
collect network traces for security research.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Katerina Papadaki |
Abstract:
This paper describes a class of patrolling games on graphs, motivated by the problem of patrolling a network vulnerable to viral infection or a facility (for example in order to defend an art gallery against theft of a painting, or an airport against terrorist attack). The network/facility can be thought of as a graph Q of interconnected nodes (e.g. rooms, terminals) and the Attacker can choose to attack any node of Q within a given time T. He requires m consecutive periods there, uninterrupted by the Patroller, to commit his nefarious act (and win). The Patroller can follow any path on the graph. Thus the patrolling game is a win-lose game, where the Value is the probability that the Patroller successfully intercepts an attack, given best play on both sides. We determine analytically optimal (minimax) patrolling strategies for various classes of graphs, and present numerical results for some intractable cases.
Joint work with Steve Alpern and Alec Morton
Short Bio:
Katerina Papadaki is a tenured Lecturer in the Operational Research Group, Department of Management at the London School of Economics. She received her PhD from Princeton University in 2002, from the Department of Operational Research and Financial Engineering, her MSc in Operational Research from the London School of Economics in 1996, and her BA in Pure Mathematics and Statistics from the University of California at Berkeley in 1994.
A major component of her research has been in developing algorithms to solve stochastic multidimensional dynamic programs that arise in dynamic resource allocation problems with applications involving physical resources (transportation networks), and radio resource allocation (wireless communication networks). Subsequently, using discrete optimization techniques she has developed algorithms for scheduling and routing problems in cellular wireless networks. Amongst others, she has worked on robust optimization techniques for scheduling, facility location routing problems, and network optimization on vehicular communications and intelligent transportation systems. Recently, her research attention has been on game theoretic problems. These include fair allocation of resources in telecommunication networks using cooperative game theory, inspection games with applications in inspections of NHS hospitals, and network patrolling games with applications in network security. She is associate editor of Optimization Letters and a member of INFORMS and IEEE.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
|
Speaker: Dr. Dimitrios Tsoumakos |
Abstract:
Peer-to-Peer (P2P) computing has gained a lot of attention from both the scientific and the large Internet user community. Popular applications utilizing this new technology offer many attractive features to a growing number of users while taking up a large amount of everyday network traffic.
This talk presents bandwidth-efficient and adaptive algorithms to facilitate data location and processing for massive data management applications that operate on P2P overlays. The basis of these schemes is their ability to learn from past interactions, increasing their performance with time.
In the first part of the talk, previous work in efficient content location and distribution for Unstructured Peer-to-Peer overlays is described. The Adaptive Probabilistic Search (APS) scheme utilizes directed walkers to forward queries on a hop-by-hop basis. Peers store success probabilities for each of their neighbors in order to efficiently route towards object holders. In the GrouPeer project, we apply many of these techniques in order to identify and group peers with similar schemas in an interconnected network of autonomous databases.
In the second part of the talk I will present some of my current work which focuses on indexing methods for data and query-intensive applications over P2P overlays. HiPPIS and PASSION are systems that utilize adaptive algorithms that automatically adjust the level of indexing (for hierarchically organized data or ranges respectively) according to the granularity of the incoming queries, without assuming any prior knowledge of the workload. Brown Dwarf is a complete system for distributing and querying data-cubes w.r.t. load and network/node failures.
Short Bio:
Dimitrios Tsoumakos is a visiting lecturer at the Computer Science Department of UCY. He received his Diploma in Electrical and Computer Engineering from NTUA in 1999, joined the graduate program in Computer Sciences at the University of Maryland in 2000, where he received his M.Sc. (2002) and Ph.D. (2006). He has been collaborating as a senior researcher with the Computing Systems Laboratory in the Department of Electrical and Computer Engineering of the National Technical University of Athens (NTUA) since 2006.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
|
Speaker: Dr. Antonis Loizou |
Abstract:
There has recently been a rapid increase in the commercial use of Recommender System technologies, primarily by online retailers. Such systems appear attractive to retailers, since they can be used to identify any products from their catalogue that can be expected to appear interesting to a particular customer, increasing the amount of purchases made. Thus, they are typically designed under assumptions that an exhaustive index of resources for recommendation is available, and that users can be adequately characterised solely through their interactions with such resources. The objective of my work is to show that by automatically and unobtrusively compiling a profile of user activities, a much more complete user representation can be obtained. Furthermore, resources for recommendation can be dynamically introduced to systems using such profiles, by importing preference data from external communities and social networks, thus enabling the provision of recommendations from multiple domains. A methodology for mapping user profiling elements, as well as the resources available for recommendation to Wikipedia articles has been developed to facilitate comparisons and address the problem of heterogeneity. Hyperlinks between Wikipedia articles are assumed to convey latent semantic relationships between the concepts they describe, and used to construct a graph using articles as nodes and hyperlinks as edges. A Markov chain model is then imposed over the graph, and exploited to drive the recommendation engine.
Short Bio:
A. Loizou received his PhD from the University of Southampton in June 2009, working at the Intelligence, Agents, Multimedia group under the supervision of Dr. Srinandan Dasmahapatra and Prof. Paul H. Lewis. He received his undergraduate degree in Computer Science with Artificial Intelligence also from the University of Southampton in 2005. With a background in Machine Learning and Probabilistic Reasoning, his PhD work has been in the field of Recommender Systems with a particular focus on developing systems that are able to provide recommendations from multiple domains. His research interests also include Semantic Web technologies, Multimedia annotation, Information Retrieval and Data Mining.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Michele Mazzucco |
Abstract:
Resource multiplexing in commercial data centers is an effective way
to achieve server consolidation, which can in turn increase the
system's throughput as well as reduce the power consumption of
Internet utilities.
In this talk I will introduce a performance management model for
monitoring the server's behavior and autonomically computing the
optimal Multi-Programming Level (MPL) accordingly, i.e., the optimal
number of jobs allowed to run concurrently, as a way to control the
scheduling of jobs. Since the MPL can heavily influence the achievable
performance but can not be computed off-line (it depends on several
factors that can't be determined statically) a framework for
autonomously controlling the MPL level is discussed. I will report the
results of several experiments that have been carried out, showing
that the proposed scheme can dynamically choose the 'best' MPL and
reduce the average response time up to 33%.
Short Bio:
Michele Mazzucco is currently a Marie Curie Fellow at the University
of Cyprus. He received his MSc from University of Bologna (Italy) in
2005, while he earned his PhD at Newcastle University (UK) in 2009
with a thesis focused on the design of adaptive service provisioning
systems subject to QoS constraints. From 2006 to 2008 he worked as a
Research Associate at Newcastle University (UK) on a BT funded
project, while since March 2008 he is one of the committers at the
Apache Software Foundation. His research interests include distributed
systems and middleware, software architectures, autonomic computing
and Quality of Service. Awards include a best paper and two EU patents.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Ioannis Anagnostopoulos |
Abstract:
In this lecture, we will introduce several metrics and techniques for dealing with the size and other evolution rates of the web. The exponential growth of the web poses a serious challenge for the Internet/web search services (publicly known as search engines), due to the fact that their effectiveness relies on their information coverage. However, web search services not only have to cover an increasing quantity of information, but also to deal with evolution incidents, since new web documents and objects are relentlessly added, old ones are moved, while others frequently have their content changed or updated. The problem of measuring such characteristics stands as a non-trivial problem due to the nature and the structure of web itself. We will present measurements derived from real experiments conducted in the well-known Internet search services such as Google and MSN.
Short Bio:
Dr Ioannis E. Anagnostopoulos was born in Athens, Greece in 1975. He received his diploma from the Department of Electrical Engineering and Computer Technology, University of Patras, Greece, in 1998 and his PhD from the School of Electrical and Computer Engineering, National Technical University of Athens (NTUA), Greece, in 2004. Currently, he is with the University of the Aegean, at the Department of Information and Communication Systems Engineering, serving as a lecturer. His research interests include Internet and Web Technologies, Search and Retrieval Software Methodologies, E-Commerce, Telecommunication Networks and Intelligent Information Systems. Dr. Ioannis Anagnostopoulos is a member of the technical chamber of Greece, IEEE, IEE and ACM.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Alexander A. Shvartsman |
Abstract:
The problem of cooperatively performing a collection of tasks in a decentralized setting where the computing medium is subject to undesirable perturbations is one of the fundamental problems in distributed computing, with applications encompassing such important areas as Internet supercomputing, parallel simulation, and multi-agent collaboration. The perturbations in the computing medium are typically due to processor and software failures (benign or malicious), communication breakdowns, and unpredictable delays. Such perturbations become even more prominent when an application needs to harness massive amounts of available computational resources. To develop efficient solutions for computation problems based on distributed cooperation, it is important to understand efficiency trade-offs characterizing the ability of p processors to cooperate on t tasks in key models of computation in the presence of adversity. In this talk we survey historical and recent results for distributed cooperation roughly grouped along the following topics: (i) fundamental failure-sensitive bounds for distributed cooperation problems for synchronous crash-prone processors, (ii) upper and lower bounds on distributed cooperation in shared-memory models, (iii) bounds on distributed work in message-passing models and on redundant work for processors that may experience prolonged absence of communication.
Short Bio:
Alexander A. Shvartsman is a Professor and Associate Head in the Department
of Computer Science and Engineering at University of Connecticut, USA. He is
the Director of the Dependable Distributed Systems Group and of the Voting
Technology Research Center. His research interests are in the Principles and
Practice of Distributed and Parallel Computing. His research has been funded by
several NSF grants, including the NSF Career Award. He has authored more
than 120 papers, two books, and several book chapters. Dr. Shvartsman has
chaired and served on many program committees of the top conferences in
Distributed Computing, he chaired the Steering Committee of DISC (2004-2007),
and he is a Vigneron d'Honneur of Jurade de Saint-Emilion.
For more info: http://www.cse.uconn.edu/~aas/
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Donald M. Chiarulli |
Abstract:
This research is aimed at the development of a new interconnection network and control
architecture for large-scale multi-core processors. It is designed to operate efficiently in systems
with hundreds to thousands of active processor cores and provides a fully interconnected
topology. Multiple programming models including symmetric common memory architectures are
directly supported without significant restrictions imposed by the underlying network. Our
specific focus is on a innovative set of design paradigms for these systems that are adaptable to
both current CMOS electronic interconnection technology as well future silicon-optics
technology.
There are two fundamental ideas behind these paradigms. First, in designing both the physical
interconnection network and the control algorithms, we endeavor to migrate complexity to the
edges of network. This means that there will be little or no intelligence or routing capability in the
network core. Instead, the physical interconnection network model is a simple many-to-many
bus-style interconnection with distributed routing and access control decisions made exclusively
the node interfaces. To make the network scalable, the strategy is to partition the network into
sub-nets with multiple transceivers, one per subnet at each node interface. Each subnet connects
the transmitters for all of the nodes in one partition to the receivers at all of the nodes in one
other.
The second fundamental idea is the basis of the scalable control architecture. Once again, our
criterion is that all routing and control decisions must be fully distributed across nodes at the
edges of the network. However, since there are no scalable solutions that can provide the global
information at the timescale of individual bus transactions, bus access at this level is governed
using a simple greedy algorithm. Each node claims a bus transaction on demand without regard to
any pending claims by other nodes. When conflicts occur, a hardware encoded fixed rule, such as
physical ordering on the bus, determines the winner. On a second level, in a time base spanning
multiple transactions, a negative feedback mechanism is used to throttle the greedy algorithm at
each node. When a node anticipates bus activity, it broadcasts a negative feedback message to all
nodes. At every active node, the amount of negative feedback present limits the level of greed. At
the time scale of this control algorithm, network bandwidth is near optimally allocated with a
small percentage reserved to allow non-active nodes to initiate. Both levels allow for a great deal
of flexibility. Overall bandwidth can allocated on a per node basis in the first level, by locally
adjusting the strength of feedback. The second level control algorithm can operate on a demand
basis or it can be made predicative by modeling software behavior or linkage to cache
management algorithms.
Short Bio:
Donald M. Chiarulli is a Professor of Computer Science and Computer Engineering at the
University of Pittsburgh. He received his M.S. degree in Computer Science from the Virginia
Polytechnic Institute and his Ph.D. also in Computer Science from Louisiana State University.
Dr. Chiarulli's research interests are in Computer Architecture and are specifically focused on the
application of novel technology to interconnection networks, system packaging, and mixed
technology integration. Contributions from Dr. Chiarulli's group have included the demonstration
of the first all-optical address decoder and several designs for time/space multiplexed data bus
architectures. Recent contributions include the Partitioned Optical Passive Star (POPS)
architecture for multiprocessor interconnection networks and the Multi Bit Differential Signaling
(MBDS) methodology. Dr. Chiarulli has authored or co-authored over 40 technical papers
including two that earned best paper awards at the International Conference on Neural Networks
(ICNN) and the Design Automation Conference (DAC) respectively. Dr. Chiarulli is a member of
the IEEE, and the SPIE.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Gaye Banfield |
Abstract:
Empirical data in psychology suggests that we recognize we have self-control problems and attempt to overcome them by exercising precommmitment, which biases our future choices to a larger, later reward. The behavioral model of self-control as an internal process is taken from psychology and implemented, using a top-down approach, as a computational model of the human brain. The higher and lower brain systems, represented by two Artificial Neural Networks (ANNs) using reinforcement learning, are viewed as cooperating for the benefit of the organism, as opposed to the classical view of the higher brain overriding the lower brain. The ANNs are implemented as two players, learning simultaneously, but independently. Psychological studies suggest that the structure of the self-control problem can be likened to the Iterated Prisoner's Dilemma game in that cooperation is to defection what self-control is to impulsiveness. I hypothesise that increasing precommitment increases the probability of cooperating with oneself in the future. To this aim, precommitment is implemented in one of three ways. The first investigates the effect of implementing precommitment by simply varying the input value of the ANN's bias node between 0 and 1 (instead of fixed as 1). This method is referred to as the 'variable bias' method. The second implements precommitment as an extra input to the ANNs in the 2-ANNs model. In this case the ANN's threshold is implemented in the usual way, i.e., as a node with an input value of 1 whose weight is trainable in the same way as the other nodes in the network and precommitment is implemented as an additional node to the input layer. This method is referred to as the 'extra input bias' method. The final method implements a bias towards future rewards as a differential bias applied to the payoff matrix. Again the ANN's threshold is implemented in the usual way, i.e., as a node with an input value of 1 whose weight is trainable in the same way as the other nodes in the network. This method is referred to as the 'differential bias method'. Finally, I investigate what role evolution has played in shaping our willingness to precommit to future rewards by subjecting the model to simulation of evolutionary adaptation. Results suggest that evolution, as opposed to learning is the key player.
Short Bio:
Dr Gaye Banfield has been involved in computing in one form or another
for twenty-six years. She was awarded a Bachelor of Science from the
University of Queensland in 1983 majoring in Computer Science with
electives in Mathematics and Psychology. From 1983 to 2002 she has been
employed in various I.T. roles and applications in manufacturing retail
and finance. Her tasks have included critical analysis of documentation,
data collection, statistical analysis and interpretation of information.
In 1998 Gaye began a Master of Science in Computer Science at Birkbeck
College part-time. She continued to work fulltime and study in the
evenings. Her MSc thesis touched on her areas of interest on AI and
Neural Networks. She graduated with a Master of Science in Computer Science in 2001
(Birkbeck, University of London). In 2006 she was awarded a PhD from Birkbeck College,
University of London for her work on computational modelling of self control. Since then
she has continued pursing her interest in Neural Networks specifically in the area of
reinforcement learning and also in using computers in mathematical education.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Eleni D. Karatza |
Abstract:
Distributed real-time systems are very important in our daily life as most of today?s applications require high quality results within precise timing constraints.
In real-time systems the correctness of the system does not depend only on the logical results of the computations, but also on the time at which the results are produced. That is, the jobs in a real-time system have deadlines which must be met. If a real-time job cannot meet its deadline, then its results will be useless, or even catastrophic. Therefore, a real-time system must guarantee that every job will complete its execution before its deadline. Moreover, it must tolerate possible software faults that may cause failures during the execution of a job.
Consequently, the most important aspect of a distributed real-time system is the scheduling algorithm which decides the allocation of processors to jobs and also the order in which jobs will be executed on processors. One of the techniques that have been proposed by researchers for the scheduling of real-time jobs is called imprecise computations. This is the case where the execution of a real-time job is allowed to return intermediate (imprecise) results of poorer, but still acceptable quality, when the deadline of the job cannot be met.
In a distributed real-time system, jobs usually consist of frequently communicating tasks which can be processed in parallel. An efficient way to schedule dynamic, parallel jobs is gang scheduling. With this technique, parallel job tasks are scheduled and executed simultaneously on different processors. Jobs of this type are called gangs.
In this talk we present issues related to the performance of scheduling algorithms for gangs in distributed real-time systems, where transient software faults may cause failures during the execution of a job. Particularly, the imprecise computations technique is discussed and the advantages of incorporating this technique into the scheduling
process are presented.
Short Bio:
Eleni D. Karatza is an Associate Professor in the Department of Informatics, at the Aristotle University of Thessaloniki, Greece. Her research interests include Computer Systems Modeling and Simulation, Performance Evaluation of Parallel and Distributed systems, Resource Allocation and Scheduling, Cluster Computing, Grid Computing and Resource Discovery in the Grid. Dr. Karatza has authored or co-authored over 130 technical papers and book chapters including two papers that earned best paper awards at the 39th Annual Simulation Symposium (ANSS 2006) and the 10th International Symposium on Performance Evaluation of Computer and Telecommunication Systems (SPECTS 2007) respectively. Dr. Karatza is Editor in Chief of Simulation Modelling Practice and Theory (Elsevier), Area Editor of the Journal of Systems and Software (Elsevier), Associate Editor of the International Journal of Communication Systems (Wiley), Associate Editor of the ACM Transactions on Modeling and Computer Simulation, Associate Editor of the International Journal of Simulation and Process Modelling (Inderscience Publishers), Editorial Board Member of the International Journal of Communication Networks and Distributed Systems (Inderscience Publishers), Editorial Board Member of the International Journal of Simulation: Systems, Science & Technology (The UK Simulation Society), Editorial Advisory Board Member in the Book Series: Emerging Communication and Service Technologies (Troubador Publishing Ltd) and Advisory Editorial Board Member of Simulation: Transactions of The Society for Modeling and Simulation International (Sage Publications). Dr. Karatza is a Senior member of the IEEE, and of the Society for Modeling and Simulation International (SCS).
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Xenofontas Dimitropoulos |
Abstract:
Contractual relationships between Autonomous Systems (AS) affect inter-domain packet routing and shape the evolution and properties of the global AS-level topology of the Internet. In this talk, I will first describe the problem of inferring AS relationships and then will introduce novel inference heuristics finding customer-to-provider, peer-to-peer, and sibling-to-sibling relationships. I will outline validation results based on a survey with network operators showing inference accuracy between 82.8% and 96.5%. Finally, I will discuss an AS relationships repository we have opened to make our results useful for the community where we archive periodically the Internet AS-level topology annotated with inferred AS relationships.
In the second part of the talk, I will switch to discussing the problem of computing network traffic heavy hitters using limited memory resources. I will briefly introduce the IBM Aurora system, which provides the context of our interest and then I will present an algorithm called Probabilistic Lossy Counting (PLC) for finding network traffic heavy hitters. PLC enhances the well-known lossy counting algorithm using on a tighter error bound on the estimated sizes of traffic flows providing probabilistic rather than deterministic guarantees on its accuracy. Performance comparison experiments show that PLC has between 34.4% and 74% lower memory consumption and between 37.9% and 40.5% fewer false positives than other state-of-the-art algorithms.
Short Bio:
Xenofontas Dimitropoulos is a Senior Researcher in the Communication Systems Group (CSG) of ETH and an Associate Tutor in the Open University of Cyprus (OUC). He received a PhD degree in Electrical and Computer Engineering from Georgia Tech. In the past, he was a post-doc in the IBM Zurich Research Laboratory, where he worked in the IBM Aurora traffic flow collector project (now part of the IBM Tivoli suite), and a visiting scholar in the Cooperative Association for Internet Data Analysis (CAIDA). His research interests focus on traffic flow measurements, inter-domain routing, and network simulation. He has had various honors, like leading the graduation oath in his BSc degree for the highest GPA, a Fulbright scholarship, a Marie Curie scholarship, a best paper award, and a best paper nomination.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science, the Department of Physics and the Department of Chemistry at the University of Cyprus cordially invite you to the colloquium entitled:
 |
Speaker: Prof. Thom H. Dunning, Jr. |
Abstract:
Advancing computational science and engineering requires progress along many axes, from the development of the underlying theories and models to the development of new algorithms and computational applications to the validation of the new theories and models. During the past two decades, chemical physicists have made dramatic advances in their ability to predict the structures, states, energetics and reactivities of molecules. However, advances are still needed: hypervalent molecules present conceptual, if not computational, difficulties and the new generation of multicore and many-core processors, especially as embodied in the coming generation of petascale computers, provide new opportunities but present new challenges as well. We will explore these issues in the seminar.
Short Bio:
Thom H. Dunning, Jr., is the director of the National Center for Supercomputing Applications as well as the Institute for Advanced Computing Applications and Technologies and a professor in the Department of Chemistry at the University of Illinois at Urbana-Champaign. Before joining Illinois, he was the founding director of the Joint Institute for Computational Sciences in Oak Ridge, Tennessee, a distinguished scientist in computing and computational sciences at Oak Ridge National Laboratory, and a distinguished professor at the University of Tennessee at Knoxville. Prior to moving to Tennessee, Dr. Dunning was responsible for supercomputing and networking for the University of North Carolina System and was a professor of chemistry at the University of North Carolina at Chapel Hill.
Preceding the above academic appointments, Dr. Dunning spent 27 years as a staff member and research director in the Department of Energy's national laboratories (Los Alamos National Laboratory, Argonne National Laboratory, and Pacific Northwest National Laboratory). He then spent two years in DOE's Office of Science as Assistant Director for Scientific Simulation, where he initiated the Scientific Discovery through Advanced Computing (SciDAC) program. Dr. Dunning received his bachelor's degree in chemistry from the Missouri University of Science & Technology (1965) and his doctorate in chemistry/chemical physics from the California Institute of Technology (1970). He has written nearly 150 scientific publications on topics ranging from advanced computational techniques for molecular calculations to computational studies of the spectroscopy of high-power lasers and the chemical reactions involved in combustion. Dr. Dunning is a fellow of the American Physical Society and the American Association for the Advancement of Science and received the E. O. Lawrence Award in 1997 and DOE's Distinguished Associate Award in 2001.
Lecture in PDF: Colloquium
|
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
| Sponsor: The CS Colloquium Series is supported by a generous donation from |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Lawrence Snyder |
Abstract:
The fastest computer in the world has achieved a speed of 10^15 floating point operations per second; all desktop and laptop computers sold today are parallel computers. What programming techniques can be used to effectively translate the potential parallelism in a computation to these kinds of computers? Will one language work for both situations?
Should all programmers be parallel programmers? The lecture discusses answers to these questions as well as other urgent problems in parallel computation.
Short Bio:
Lawrence Snyder is a professor of Computer Science and Engineering at the University of Washington in Seattle. He received a BA from the University of Iowa in Mathematics and Economics, and his PhD from Carnegie Mellon University as a student of A. Nico Habermann. He has served on the faculties of Yale and Purdue, and has had visiting appointments at UW, Harvard, MIT, Sydney University, The Swiss Technological University (ETH), The University of Auckland and Kyoto University. Throughout most of his career Snyder's research has focused on parallel computation, including architecture, algorithms and languages. In 1980 he invented programmable interconnect, a method to dynamically configure on-chip components, and a technology used today for FPGAs. In 1990 he was co-designer of Chaos Router, a randomizing adaptive packet router. He was principle investigator of the ZPL language design project, the first high-level parallel language to achieve "performance portability" across all parallel computer platforms. Snyder is author of Fluency with Information Technology: Skills, Concepts and Capabilities, a textbook for non-techie college freshmen that teaches fundamental computing concepts; the book is in its 3rd edition. With former PhD student Calvin Lin (UT Austin), he has written Principles of Parallel Programming, published in 2008. In service, Snyder was a three-term member of the Computer Research Association Board of Directors, developing a series of best practices white papers. He chaired the NSF CISE Advisory Board as well as several CISE directorate oversight panels and numerous review panels. He has chaired two National Research Council studies, producing influential reports -- Academic Careers for Experimental Computer Scientists and Engineers and Being Fluent with Information Technology; he served three terms on NRC's Army Research Lab Technical Advisory Board. He serves on ACM's Education Board, has been general chair or program committee chair of several ACM and IEEE conferences. He is a fellow of both the ACM and IEEE. His most important and rewarding accomplishment has been as adviser to 21 doctoral students.
Lawrence Snyder will be a short term visiting professor at the Department of Computer Science, University of Cyprus, during Spring 2008-2009.
|
Additional Lectures by Prof. Snyder:
For additional information please download the following: Lectures Overview in PDF |
|
Lectures in PDF: Colloquium,
Lecture 1,
Lecture 2
|
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Thanasis Hadzilacos |
Abstract:
What is m-learning about? Is it about delivering information on smaller screens via lesser bandwidth? Is it about giving just-on-time and just-acceptable education to those unfortunate ones outside a classroom, distant from their classmates and a 'real' teacher? We shall argue in this presentation that hardware, software and communication properties and restrictions -valid and important computer science and technology research issues as they may be- are incidental to the deeper problems and opportunities that m-learning presents. We shall argue that m-learning provides an opportunity to bridge a classic gap in education, that between "library learning" and "field learning", a gap as old as the written word.
"Content" is to education as "instruments" is to music: essential, indispensable, but a far cry from being the whole. Learning is not information presentation. An architecture for m-learning should be conceived and designed around complex learners' educational activities and not around content browsing -a vital but very simple learning activity. We shall discuss architectural issues and propose an architecture based on constructive m-learning activities.
Is context awareness a desirable characteristic, a necessary one, something already achieved, or a pie in the sky? Is 'context' a stand-alone concept or is it context dependent? Does 'context awareness' for m-learning systems simply mean 'learner location dependence'? We shall argue that there are indeed m-learning system architectural implications of context awareness which depend on the answer we give to such questions.
Short Bio:
Professor of Information Systems, Open University of Cyprus (http://www.ouc.ac.cy) academic director of the graduate program in Information Systems, 55 years old. Until September 2007 he was Dean of the School of Science and Technology at the Hellenic Open Univer-sity (HOU, http://www.eap.gr) where he served (2000-2007) as associate professor of Software Engineering, directed (2003-2007) the Open and Distance Laboratory for Educational Material and Educational Methodology, the graduate course on Information Systems (2003-2007) and the undergraduate Computer Science course (2001-2003). Educated at Harvard, USA (1971-76), he had substantial industrial experience before joining Computer Technology Institute (http://www.cti.gr) in 1986, where he continues as a researcher with the responsibility of the Educational Technology and the e-Learning Sectors and R&D Unit III "Applied Information Systems" (http://www.cti.gr/RD3). He has taught at the Universities of Patras and Thessaly before joining the Hellenic Open University in 2000. During 1996-2001 he designed and managed the Greek national project "Odysseia" (http://odysseia.cti.gr/) for the utilization of Information and Communication Technologies in secondary education. He has served as a member of the Council of Europe working group for Teaching and Learning in the Communication Society (2002-2004) and the Greek national representative to E.U. DG Education and Culture for building the European portal on educational opportunities (2002-2005). He has published over 80 papers in international journals and conferences, including a chapter on "Teaching and Learning in the Communication Society" published by the Council of Europe. He has given over 60 invited talks and presentations in scientific conferences, training seminars, university seminars, and professional or technical events. He has coordinated, directed and participated in over 40 research and development projects funded by the European Commission (IST, Esprit, Brite-Euram, eContentplus, Lingua, Minerva, e-Learning), the three Community Support Framework Programs for Greece, private companies, the Greek Secretariat for R&D and the Greek Ministry of Education. His research interests are related to education and to large-scale information and database systems and in particular system design for non-standard application areas such as education, GIS, and multimedia. His real interest is people, and he is currently studying theology at HOU.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Christos Kapoutsis |
Abstract:
Understanding the power of nondeterminism is one of the major goals of the theory of computation. The most important problem in this respect is the famous P vs NP question: does nondeterminism make a difference on Turing machines that use only "small" (i.e., polynomial) time? Another important problem is the L vs NL question: does nondeterminism make a difference on Turing machines that use only "small" (i.e., logarithmic) space?
In 1978, Sakoda and Sipser proposed the following analogue to these questions: instead of full-fledged Turing machines, focus only on those which cannot write on their tape; instead of time or space, focus on size. That is: does nondeterminism make a difference on two-way finite automata that use only a "small" (i.e., polynomial) number of states?
Also known as the 2D vs 2N question, where 2D (resp., 2N) is the class of problems that can be solved by small deterministic (resp., nondeterministic) two-way finite automata, this problem remains open. The conjecture is that indeed 2D and 2N are different. Given that 2D is closed under complement, one way to confirm the conjecture is to show that this closure fails for 2N; namely, that complementing a nondeterministic two-way finite automaton involves an exponential blow-up in the number of states, in general. In this colloquium, we will sketch a proof of this claim for the special case of automata that are sweeping, in the sense that they can change the direction of their head only at the two ends of the tape.
Short Bio:
Christos Kapoutsis began his graduate studies at MPLA, Athens and continued to receive his PhD from the MIT EECS Department in 2006, for work on the size complexity of finite automata. After two years as a postdoctoral researcher at the Chair for Information Technology and Education at ETH, he is now a visiting lecturer at the Department of Computer Science, University of Cyprus.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Andries P. Engelbrecht |
Abstract:
The original particle swarm optimization (PSO) algorithms have been developed to solve unconstraint, static continuous-valued optimization problems. Due to the characteristics of PSO, it cannot be applied to find solutions in dynamically changing environments. The PSO approach has to be adapted in order to inject diversity into swarms such that the exploration abilities of the swarm are increased. This then allows PSO to find and track optima in dynamic environments. This talk will start by formally defining dynamic environments and discussing different classes of dynamic environments, as well as classes of dynamic optimization problems. Then an introduction to PSO will be provided, with an explanation of why the original PSO cannot be used in dynamic environments. Adaptations of PSO to find and track single solutions in dynamic, single objective, and unconstrained environments will then be discussed. The talk will then continue to discuss more complex dynamic optimization problems. It will be shown how PSO can be adapted to track multiple solutions in a dynamic environment, and results will be given to illustrate the performance of PSO in this task. Dynamic multi-objective optimization problems will be considered, discussing how a vector-evaluated PSO can be used to solve dynamic multi-objective optimization problems. Finally, the ability of PSO to cluster temporal data will be illustrated.
Short Bio:
Andries Engelbrecht is a professor in Computer Science at the University of Pretoria, South Africa. He also holds the position as South African Research Chair in Artificial Intelligence, and leads the Computational Intelligence Research Group at the University of Pretoria, consisting of 50 Masters and PhD students. He obtained his Masters and PhD degrees in Computer Science from the University of Pretoria in 1994 and 1999 respectively. His research interests include swarm intelligence, evolutionary computation, artificial neural networks, artificial immune systems, and the application of these CI paradigms to data mining, games, bioinformatics, and finance. He has published over 130 papers in these fields in journals and international conference proceedings, and is the author of the two books, "Computational Intelligence: An Introduction" and "Fundamentals of Computational Swarm Intelligence". In addition to these, he is a co-editor of the upcoming books, "Applied Swarm Intelligence" and "Foundations on Computational Intelligence". He is very active in the international community, annually serving as a reviewer for over 20 journals and 10 conferences. He is an associate-editor of the IEEE Transactions on Evolutionary Computation, Journal of Swarm Intelligence, and the recent IEEE Transactions on Computational Intelligence and AI in Games. Additionally, he serves on the editorial board of 3 other international journals, and is co-guest-editor of special issues of the IEEE Transactions on Evolutionary Computation and the Journal of Swarm Intelligence. He served on the international program committee and organizing committee of a number of conferences, organized special sessions, presented tutorials, and took part in panel discussions. As member of the IEEE CIS, he is a member of the Games technical committee and chair of its Swarm Intelligence for Games task force. He also serves as a member of the Computational Intelligence and Machine Learning Virtual Infrastructure Network.
Andries Engelbrecht will be a short term visiting professor at the Department of Computer Science, University of Cyprus, during Spring 2008-2009.
|
Additional Lectures by Prof. Engelbrecht:
For additional information please download the following: Lectures Overview in PDF |
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Chrisina Draganova |
Abstract:
In this presentation we discuss the use of Neural Network based methods for three applications related to processing facial information. The first application aims to design neural network based classifiers that accept low dimensional representations of unseen images and produce an estimate of the age of the person in the corresponding face images. Supervised and unsupervised neural networks are tested. The results are compared with results obtained with other existing classifiers. Finally a comparison with human performance in the task of age estimation is presented.
The second application investigates the use of neural network based methods for learning the relationship between certain facial attributes and a coded representation of face images. We then use the resulting neural networks for the synthesis of face images with specific attributes.
The third application addresses the problem of restoring the overall shape of faces given only the shape presentation of a small part of the face. The shape of a face is defined by a series of landmarks located on the face outline and on the outline of different facial features. We use of a number of methods including a method that utilizes a Hopfield neural network, a method that uses Multi-Layer Perceptron (MLP) neural network, a novel technique which combines Hopfield and MLP together, and a method based on associative search. These techniques could form the basis for developing face image processing systems capable of dealing with occluded faces.
Short Bio:
Chrisina has an MSc degree in Computing Science from Birkbeck College, University of London, an MSc degree in Mathematics and Informatics, and a Ph.D. degree in Applied Mathematics from Sofia University, Bulgaria. She has worked in a number of universities in the UK and Bulgaria including London Metropolitan University, South-Bank University, Kingston University, University College London and Veliko Turnovo University. At present, she is a Senior Lecturer at University of East London. Her research interests over the years have been in the areas of approximation and interpolation with spline functions, neural network applications and Internet Technologies.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Kevin Lee |
Abstract:
Scientific workflow execution on distributed grid-based resources suffers from non-optimal execution due to the use of static upfront mapping and scheduling. In this talk we argue that adaptively mapping and scheduling workflows offers increased performance and better utilisation of grid resources. We Introduce a generic architecture we have built that can retrofit adaptive behaviour to previously non-adaptive systems. Adaptation is controlled through the use of stream query processing which aggregates and performs analysis of sensor data collected from the workflow execution. New schedules are generated whilst the workflows are executing using utility function optimisation which results in improved performance. The approach is evaluated with the Pegasus workflow management system which compiles, schedules and executes abstract workflows on grid resources. Multiple utility functions are compared to optimise for response time and profit based metrics for different numbers of workflows.
Short Bio:
Kevin Lee is a Postdoc Research Associate at the School of Computer Science, University of Manchester, UK. He received his PhD on Dynamic Programming Models for Network Processors from Lancaster University, UK in 2006. His research interests include grid-based workflow execution, adaptive systems and P2P network monitoring. He is currently working on optimising scienfic workflow execution on grid infrastructure through the use of stream query processing and the optimisation of utility functions.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Manolis Christodoulakis |
Abstract:
The process of digitising old books and manuscripts is of immense importance to a variety of people, such as librarians, academics, publishers, etc. This task is achieved by scanning the documents and then performing Optical Character Recognition (OCR) to obtain text that can be stored, searched for, indexed etc. Quite often the original paper-copies of the publications are of poor print quality, leading to digital texts that contain errors. Consequently, any attempt for exact pattern matching will fail, and algorithms for approximate pattern matching must be used, where matches similar (rather than identical) to the pattern can be identified. There exist several different ways for defining text similarity, which however fail to incorporate the specific nature of errors that occur in OCR-texts. In this talk I will present a recently developed similarity measure that is specifically tailored for this purpose. In particular, it incorporates optical similarities of characters as well as matching combinations of characters to yield better approximate matching. Early implementations suggest that it is a promising method, and there is number of variants worth exploring in the future.
Short Bio:
Dr. Manolis Christodoulakis received his BSc from the Department of Computer Engineering and Informatics, University of Patras, and his PhD from the Department of Computer Science in King's College London. In the past, he has worked as a Research Associate and later as an External Lecturer in King's College. Since September 2007, he serves as a Lecturer in the Secure Systems and Software Development (SD) field, in the School of Computing, Information Technology & Engineering. His research interests include: design and analysis of combinatorial algorithms, sequence analysis (pattern matching, repetition finding etc.), computational biology/bioinformatics, and computational music analysis.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Dr. Rizos Sakellariou |
Abstract:
The talk will argue for the need to use autonomic computing techniques to improve performance in volatile environments such as those typically associated with Grids and large-scale distributed computing. The realization of autonomic computing is based on having a feedback loop that monitors the environment, makes assessments and acts if needed. Different approaches to implement this feedback loop, ranging from control theory to utility functions, will be discussed and illustrated with examples drawn from research in the context of a UK project.
Short Bio:
Rizos Sakellariou is a Senior Lecturer in the School of Computer Science of the University of Manchester. Before, he was with the Department of Computer Science, Rice University, where he was a member of the dHPF Compiler Group and was involved with the DARPA-funded project POEMS. He was also a Visiting Assistant Professor with the Department of Computer Science, University of Cyprus (Fall 1999), and also held visiting positions with the Department of Computer Science, University of Illinois at Urbana-Champaign (Summer 2000) and the Department of Computer Architecture, Universitat Politecnica de Catalunya (Fall 2002). His research interests fall in the area of High-Performance Parallel, Distributed and Grid Computing. For more information please visit: http://www.cs.man.ac.uk/~rizos/
|
Lecture in PDF: Colloquium
|
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Vasilis A. Vassalos |
Abstract:
Adaptive join algorithms have recently attracted a lot of attention in emerging applications where data is provided by autonomous data sources through heterogeneous network environments. Their main advantage over traditional join techniques is that they can start producing join results as soon as the first input tuples are available, thus improving pipelining by smoothing join result production and by masking source or network delays. I will describe Double Index NEstedloops Reactive join (DINER), a new adaptive join algorithm for result rate maximization. DINER combines two key elements: an intuitive flushing policy that aims to increase the productivity of in-memory tuples in producing results during the online phase of the join, and a novel re-entrant join technique that allows the algorithm to rapidly switch between processing in-memory and disk-resident tuples, thus better exploiting temporary delays when new data is not available. I will present experimental results using real and synthetic data sets that show that DINER outperforms previous adaptive join algorithms in producing result tuples at a significantly higher rate, while making better use of the available memory.
Short Bio:
Prof. Vasilis Vassalos (PhD in CS, Stanford University, 2000) is an Associate
Professor in the Department of Informatics at AUEB. His research is on
infrastructure and algorithmic issues for the integration of data and Web
services in different environments, including XML query processing,
specification-driven interface generation, adaptive query processing,
and query rewriting using views. He is also working on sensor data management.
Vassalos is the recipient of numerous awards, including a Marie Curie
Outgoing International Fellowship for 2007-2008, and has been Principal
Investigator for 8 funded research and advanced development projects since
his arrival at AUEB in 2004. He is the author of over 25 technical
publications and two US patents. He is or has been a member of the
Program Committees of numerous international conferences, including SIGMOD
2008 and VLDB 2007. He is the co-founder of software company Enosys Software
(sold to BEA Systems in 2003), maker of the first XQuery-based data
integration platform and XQuery engine. Before joining AUEB he was an
Assistant Professor of Information Systems at the Stern School of Business
at NYU.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Lawrence Rauchwerger |
Abstract:
Hybrid Analysis (HA) is a compiler technology that can seamlessly integrate all static and run-time analysis of memory references into a single framework capable of generating sufficient information for most memory related optimizations.
In this talk, we will present Hybrid Analysis as a framework to perform automatic parallelization of loops. For the cases when static analysis does not give conclusive results, we extract sufficient conditions which are then evaluated dynamically and can (in)validate the parallel execution of loops. The HA framework has been fully implemented in the Polaris compiler and has parallelized 22 benchmark codes with 99% coverage and speedups superior to the Intel Ifort compiler.
Short Bio:
Lawrence Rauchwerger is a Professor Computer Science and of Computer Engineering in the Department of Computer Science, Texas A&M University. He is also the co-Director of the Parasol Laboratory. He received an Engineer degree from the Polytechnic Institute Bucharest, a M.S. in Electrical Engineering from Stanford University and a Ph.D. in Computer Science from the University of Illinois at Urbana-Champaign. Since 1996 he has been on the faculty of the Department of Computer Science at Texas A&M where he co-founded the Parasol Lab. He has held Visiting Faculty positions at the University of Illinois at Urbana-Champaign, Bell Labs, IBM T.J. Watson Research Center, and INRIA FUTURS, Paris.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Manolis Katevenis |
Abstract:
Communication (both interprocessor and I/O), i.e. data movement, is at least as important as computation, especially in multiprocessors.
To reduce latency, new architectures will need to bring the supporting hardware, i.e. the network interfaces (NI), close to each processor, hence at the same level as their caches. Both the cache controller and the NI move data, thus they can benefit from being merged together.
Implicit communication occurs when we do not know in advance which input data will be needed, or who last modified them; cache coherence works well for such communication. Explicit communication is when the producer knows who the consumers will be, or when the consumer knows its input data set ahead of time. Cache prefetchers or remote DMA (RDMA) are effective transfer mechanisms for explicit communication; however, RDMA uses 3 to 5 times less packets for an equivalent transfer, thus saving a lot of energy. In the SARC project, we are designing CMP nodes where the local SRAM blocks of the processor are configurable as partly-cache and partly-scratchpad memory, and where the cache controller and network interface are merged together, thus unifying the hardware support for implicit and explicit communication.
Short Bio:
Manolis Katevenis received the Ph.D. degree from U.C.Berkeley in
1983 and the ACM Doctoral Dissertation Award in 1984 for his thesis on "Reduced Instruction Set Computer Architectures for VLSI".
After a brief term on the faculty of Computer Science at Stanford University, he is in Greece, with the University of Crete and with FORTH, since 1986. After RISC, his research has been on interconnection networks and interprocessor communication.
In packet switch architectures, his contributions since 1987 have been mostly in per-flow queueing, credit-based flow control, congestion management, weighted round-robin scheduling, buffered crossbars, and non-blocking switching fabrics. In multiprocessing and clustering, his contributions since 1993 have been on remote-write-based, protected, user-level communication.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
The Department of Computer Science at the University of Cyprus cordially invites you to the colloquium entitled:
 |
Speaker: Prof. Babak Falsafi |
Abstract:
Device scaling in processor fabrication technologies along with microarchitectural innovation have led to a tremendous gap between processor and memory performance. While architects have primarily relied on deeper cache hierarchies to reduce this performance gap, the limited capacity in higher cache levels and simple data placement/eviction policies have resulted in diminishing returns for commercial workloads with large memory footprints and adverse access patterns. Moreover, proposals to bridge the gap using runahead execution or large instruction windows do not benefit workloads with little inherent memory-level parallelism such as transaction processing on databases or web servers.
The STeMS (Spatio-Temporal Memory Streaming) project at EPFL is exploring memory system designs that exploit repetitive spatial and temporal correlation among memory accesses and construct memory streams that can be moved and managed together through the memory hierarchy to hide the long access latencies. In this talk, I will present: (a) results from offline trace analysis and cycle-accurate simulation showing that a large fraction of memory accesses in server workloads are spatially and/or temporally correlated, and (b) candidate STeMS architectures to exploit such correlation.
Short Bio:
Babak Falsafi is a Professor in the School of Computer and Communication Sciences at EPFL, and an Adjunct Professor of Electrical and Computer Engineering and Computer Science at Carnegie Mellon. He is the Microarchitecture thrust leader for the FCRP Center for Circuit and System Solutions and directs the Parallel Systems Architecture Laboratory (PARSA) at EPFL. His research targets architectural support for parallel programming, resilient systems, architectures to break the memory wall, and analytic and simulation tools for computer system performance evaluation. In 1999, in collaboration with T. N. Vijaykumar he showed for the first time that multiprocessors do not need relaxed memory consistency models to achieve high performance. He is a recipient of an NSF CAREER award in 2000, IBM Faculty Partnership Awards between 2001 and 2004, and an Alfred P. Sloan Research Fellowship in 2004. He is a senior member of IEEE and ACM.
| Web: http://www.cs.ucy.ac.cy/colloquium/ |
| Mailing List: http://listserv.cs.ucy.ac.cy/mailman/listinfo/cs-colloquium |
| RSS: http://www.cs.ucy.ac.cy/colloquium/rss.xml |
)
