2007 seminar series

Biography:

Kate Smith-Miles is a Professor and Head of the School of Engineering and Information Technology at Deakin University in Australia. Prior to joining Deakin University in 2006, she was a Professor in the Faculty of Information Technology at Monash University, Australia, where she was co-Director of the Monash Data Mining Centre. She obtained a B.Sc(Hons) in Mathematics and a Ph.D. in Electrical Engineering, both from the University of Melbourne, Australia. Kate has published 2 books on neural networks and data mining applications, and over 170 refereed journal and international conference papers in the areas of neural networks, combinatorial optimization, intelligent systems and data mining. She has been awarded over AUD$1.5 million in competitive grants, including 7 Australian Research Council grants as well as industry awards. She serves on the editorial board of several international journals, and is Chair of the IEEE Computational Intelligence Society's Technical Committee on Data Mining. In addition to her academic activities, she also regularly acts as a consultant to industry in the areas of data mining and intelligent systems.

Abstract:

Biography:

Prof Abdul Sattar is Director of the Institute for Integrated and Intelligent Systems (IIIS), a research centre of excellence at Griffith University. He has been at Griffith University since 1992 as a Lecturer (1992-95), Senior Lecturer (1996-99), and Professor (2000-present) within the School of Information and Communication Technologies. He holds a MSC in Physics (University of Rajasthan, India), MPhil in Computer and Systems Sciences (Jawaharlal Nehru University, India), and MMath in Computer Science at University of Waterloo, Canada, and a PhD in Computer Science (with specialization in Artificial Intelligence) from the University of Alberta, Canada. Professor Sattar’s current research interests include knowledge representation and reasoning, constraint satisfaction, rational agents, and propositional Satisfiability. He has supervised over 12 successful completions of PhD degrees, and published over 100 papers in conferences and journals.
His research team has won three major international awards (Gold Medal for SAT 2005 competition on random problems, IJCAI 2007 Distinguished paper award, and Gold Medal for SAT 2007 competition on random problems) in recent years. Prof Sattar is currently leading the SAFE Agents work package within NICTA’s SAFE project.

Prof. Sattar is a professional member of the American Association of Artificial Intelligence and Association of Computing Machinery.He has been actively involved in organizations/steering committees of several national and international conferences for the last 10 years.

Abstract:

The problem of finding a consistent truth assignment to all propositional variables in a formula, known as SAT problem, has been an interesting and difficult challenge. Indeed, SAT is at the heart of all computationally intractable problems. Many real world problems could be encoded as SAT problems. Thus finding an efficient solution for SAT has far reaching impact on computationally hard problems. This talk will begin with an overview of the main approaches for solving SAT problems. We will then focus on stochastic local search based methods. These methods have been shown to be highly effective for large size problems. We will present our recent results on clause weighting based local search, including an influential method that automatically learns about the structure of the problem, and efficiently exploit those structures to solve some of the difficult challenge problems.

Biography:

Shlomo Berkovsky is a Research Fellow at the Computer Science and Software Engineering Department at the University of Melbourne, Australia. His current research project focuses on Personalization of Museum Visits using Positioning and Natural Language Processing technologies. His research interests include ubiquitous and distributed user modeling, recommender systems and context-aware personalization. In the past, he graduated from the University of Haifa, Israel, where his research focused on Mediation of User Models in Recommender Systems.

Abstract:

Provision of personalized recommendations to users requires accurate modeling of their interests and needs. This paper proposes a general framework and specific methodologies for enhancing the accuracy of user modeling in recommender systems by importing and integrating data collected by other recommender systems. Such a process is defined as user models mediation. The paper discusses the details of such a generic user modeling mediation framework. It provides a generic user modeling data representation model, demonstrates its compatibility with existing recommendation techniques, and discusses the general steps of the mediation. Specifically, four major types of mediation are presented: cross-user, cross-item, cross-context, and cross-representation. Finally, the paper reports the application of the mediation framework and illustrates it with practical mediation scenarios. Evaluations of these scenarios demonstrate the potential benefits of user modeling data mediation, as it improves the quality of the recommendations provided to the users.

Biography:

Sid has been with the Monash University, Clayton Campus since September 1990. He joined the Department of Computer Science, Monash University as a Lecturer and, subsequently, was promoted to a Senior Lecturer position.

He received his PhD degree in Electrical Engineering from the Imperial College of Science, Technology, and Medicine, University of London, in 1985. He is a recipient of the Commonwealth PhD Scholarship, UK and the Fulbright Postdoctoral Scholarship, USA. He has been a Senior Member of the IEEE since 1989.

His research interests lie mainly in the areas of Pattern Recognition and Image Processing. In these areas he has published about eighty research papers and supervised over six completed PhD and Research Masters theses.

Abstract:

Content-Based Image Retrieval (CBIR), also known as Content-Based Visual Information Retrieval (CBVIR), is the process of retrieval of digital images from an image database based on the visual contents of images. For this talk we assume the “query by example” paradigm of CBIR. In this paradigm the user inputs one or more query images to the CBIR system and the system returns a number of images from the database having visual contents similar to that of the query image(s). Computationally, the visual contents of images are often represented by low level features such as colour, texture, and shape. Often this leads to the so called semantic gap, the gap between the human judgement about the visual similarity of images and the similarity obtained based on computed features.

The purpose of the talk is to discuss about a number of methods, developed in an effort to reduce the semantic gap, that have been investigated as a part of the recently completed PhD thesis of Mrs Gita Das. The following methods will be investigated: 1. a new feature representation using colour co-occurrence matrix in HSV space, 2. relevance feedback using feature re-weighting, and 3. instance-based approach using the concept of cluster density.

Biography:

Michael Kaufmann received his Diploma degree in 1983 and his Doctoral degree in 1987 from the Saarland University in Saarbrücken. In 1992 he got a professorship in Passau, and since 1993, he is professor for Computer Science in Tübingen. His interests are mostly algorithmicly oriented and range from applications in geometry, bioinformatics, and parallel computing to pure graph theory. Michael Kaufmann is a member of the editorial board of the Journal of Graph Algorithms and Applications as well as of the Journal of Discrete Algorithms and a frequently member of program committees in the field of algorithmics. He is a co-founder of the company yWorks GmbH.

Abstract:

The problem how to place appropriate labels for the different sites in maps is an important task in the area of geographical information systems, computational geometry and graph drawing. In this talk, I will review a project that has been done in the last years together with the group of Antonis Symvonis from Athens, Greece. It provides a nice example for the algorithmic approach on problem solving. We develop different models for a specific subproblem called boundary labeling. In boundary labeling, the labels are placed at the boundary of the map, and they are connected to the corresponding sites by polygonal lines, called leaders. A multiple of different algorithmic approaches ranging from very simple to quite sophisticated can be applied here.

Biography:

Barry Richardson has a BA in Psychology from Queens University, Canada; an MsC in Experimental Psychology from theUniversity of Sussex and a PhD inPsychology also from Queens University, Canada. He has worked in Bermuda, Canada, Papua New Guinea, UKand Australia. His main areas of interest are sensory processes, virtual reality and, recently, machine intelligence.

Abstract:

Some theorists (e.g., enactivist) argue that intelligence may emerge from sensory-motor links in a manner that can be applied to machines as well as humans. Specifically, a machine might learn to be adaptive if it has a way of usefully recording the consequences of what it did (motor output) as a response to a stimulus (sensory input). This action-based model implies a central role for haptics in the linking process, and for kinaesthesis in particular because it registers movement. Unfortunately, we don't know enough about how kineasthesis and touch work together (or individually) to comprise the haptic senses. This paper describes some experiments that address this problem.

Biography:

Zbigniew Michalewicz is Professor at the School of Computer Science, University of Adelaide. He completed his MSc degree at Technical University of Warsaw in 1974 and he received PhD degree from Institute of Computer Science, Polish Academy of Sciences, in 1981. He holds Doctor of Science (Habilitation) degree in Computer Science from the Polish Academy of Science (1997). From 1988 to 2004 he was Professor at University of NorthCarolina at Charlotte (USA). Zbigniew Michalewicz holds also Professor positions at the Institute of Computer Science, Polish Academy of Sciences, at the Polish-Japanese Institute of Information Technology, and a honorary Professor position at State Key Laboratory of Software Engineering of Wuhan University, China. He is also associated with Structural Complexity Laboratory at Seoul National University, South Korea. His current research interests are in the field of modern heuristic methods. He has published several books, including a monograph Genetic Algorithms + Data Structures = Evolution Programs and over 200 technical papers in journals and conference proceedings. He also published a text on modern heuristic methods, How to Solve It: Modern Heuristics, which is a standard text on hundreds of universities all over the world. The newest books include Winning Credibility: A guide for building a business from rags to riches and Adaptive Business Intelligence.

http://www.cs.adelaide.edu.au/~zbyszek

Abstract:

What is missing in most curricula - starting from elementary school all the way through to university education - is coursework focused on the development of problem-solving skills. Most students never learn how to think about solving problems in general - throughout their education, they are constrained to concentrate on textbook questions at the back of each chapter. So, without much thinking, they apply the material from each chapter to solve a few problems given at the end of each chapter (why else would a problem be at the end of the chapter?). With this type of approach to “problem solving,” it is unsurprising that students are ill prepared for framing and addressing real-world problems. When they finally enter the real world, they suddenly find that problems do not come with instructions or guidebooks. To address this gap in the educational curriculum, I have created a new 1-level course that is aimed at getting students to think about how to frame and solve unstructured problems (those that are not encountered at the end of some textbook chapter ...). The idea is to increase the student's mathematical awareness and problem solving skills by discussing a variety of puzzles. The seminar presents the case for such a course and encourages participants to introduce it in their educational units.

Presentation slides (ppt, 1.1mb)
Video recording of this seminar (mov, please note: due to the large file size we recommend you download before playing this)

Biography:

Paul Lajbcygier (BSc, MFin, PhD, GradCertHE, MACS) joined the Departments of Econometrics & Business Statistics and Accounting and Finance as an Associate Professor in 2007. Prior to that he worked in the Faculty of IT, Monash University. He combines extensive industry and academic experience in investments. Since 1990, Paul has provided investment advice for various prominent domestic and international: funds managers, banks and hedge funds. Since 1995, Paul has published over 50 academic papers and generated over $3.1 million in government grants and payments in-kind. He has sat on over 10 journal editorial boards and conference program committees. He has also worked/researched at amongst the best business schools in the world, including: London Business School and the Stern School of Business, New York University.

Abstract:

Assets under management are an essential component of a fund's overall value. We estimate a model that links asset growth to performance characteristics. We use the model to isolate significant performance characteristics and confirm that the model has predictive power out-of-sample. This model may be useful to academics, fund managers and fund allocators.

Presentation slides (ppt, 124kb)

Biography:

Ronald Pose completed his B.Sc. (hons) and his Ph.D. at Monash University, His Ph.D. involved the design and implementation of a novel capability-based operating system kernel, "Password-Capability System", and the design and construction of tightly-coupled multiprocessor hardware with novel addressing mechanisms to support it.In 1987 Ronald Pose was employed as a Research Scientist at Telecom Australia Research Laboratories where he worked on the application of public key cryptography and authentication and certification techniques. He joined the faculty of Monash University in 1988. There he has supervised a number of research students with whom he has worked on a wide variety of research projects including neural networks, genetic algorithm function optimization, network routing, low latency virtual reality address recalculation pipeline display system, self-reconfigurable computer systems. Dr. Pose's current research interests include virtual reality and telerobotics technology, computer architecture, parallel and distributed computer systems architecture, secure operating systems, reconfigurable computer systems architecture, multiprocessor interconnection networks, wireless ad hoc networks and spread-spectrum microwave communication technology, computer system security. He currently has Ph.D. students working on wireless ad-hoc networking.

Abstract:

A grand challenge of computing was to produce a system in which a person is immersed in a computer generated virtual world and can believe and behave appropriately in that world. To achieve that we have to replace the usual sensory inputs with computer generated ones. A fundamental problem in doing so is to get high quality, photo-realistic images in front of the user's eyes quickly enough. In this seminar I present a research methodology that provided new insight into the problem, overturning the conventional wisdom, and outline a novel path to its solution, drawing on ideas dating back to the ancient Greeks. With the new conceptual basis in place, a rigorous system-based engineering approach was employed to design and build a working prototype that demonstrates the efficacy of the new concepts. Aspects of computer science, computer systems engineering, software engineering, electronics, perceptual psychology, human-computer interaction, and even network engineering have been applied to create the system and investigate its applications.

Biography:

http://www.csse.monash.edu.au/~berndm/ Bernd Meyer is an Associate Professor at Monash University in Melbourne. Bernd's primary research interests are natural and nature-inspired optimization. In natural optimization he is attempting to decipher how behavioural mechanisms in biological systems lead to "optimal" decision making. He is specifically working on mathematical models of social insect behaviour and he is using insights from these models in the design of nature-inspired optimization algorithms. Having an interest in practical applications, he does however not shy away from exploiting very unnatural methods, such as constraint solving or integer programming. Bernd's second research interests are smart visual interfaces, in particular automatic interpretation of sketched diagrams, automatic diagram layout, and gesture recognition.

Abstract:

Stochastic meta-heuristics are widely used for combinatorial optimization in industrial and scientific applications. A long-standing problem with these methods has been how to handle hard constraints effectively. Generally, this entails the design and implementation of problem- specific repair methods, feasibility maintaining neighbourhood functions etc. This step requires significant expertise and is usually performed with ad-hoc methods, a problem which in practice often results in an inadequate "off-the-shelf" application of meta-heuristics. Hybridization with constraint programming, an exact technique for solving hard constraints, promises a new solution for constraint optimization with stochastic meta-heuristics. The talk discusses how such an integration can be achieved and explores different forms of couplings. An empirical evaluation demonstrates performance advantages of the hybrid algorithms for problems of intermediate tightness. An important aspect is that the proposed hybrid algorithms are merely some instances of a general framework for the integration of constraint programming with model-based constructive meta-heuristics. The talk will sketch the extension of a general framework to the wider context of Estimation of Distribution Algorithms. From a conceptual perspective, arguably the most important aspect of this approach is that it introduces declarative problem models into meta-heuristics. This opens a way to fine-tune their function for a new problem domain in a systematic way. It constitutes an important step towards greater ease-of-use of stochastic meta-heuristics, thus reducing the risk of inadequate "off-the-shelf" application.

Biography:

Korb is a Reader in the Clayton School of Information Technology, Monash University, Australia. He has always combined interests in computer science and philosophy. In the 1980s he was datacomm programmer in Santa Clara (Sillycon) Valley; before that he helped develop a time sharing operating ystem. His PhD in philosophy of science (Indiana, 1992) investigated the philosophical and computational foundations of scientific induction.He has been with Computer Science at Monash University since then. His current research continues the joint interest in machine learning and the philosophy of science, especially emphasizing the automation of causal discovery (learning Bayesian networks). He currently leads research projects on: evolutionary ALife simulations; applications of Bayesian network modeling to meteorology, ecological systems, epidemiology, poker, etc.; the causal interpretation of Bayesian nets; the theory of machine learning evaluation; informal logic and argumentation.He is co-author (with Ann Nicholson) of "Bayesian Artificial Intelligence" Chapman Hall/CRC (2004).He was a co-founder of "Psyche: An Interdisciplinary Journal of Research on consciousness".

Abstract:

In recent decades computer simulation has spread from an esoteric concern of meteorologists and econometricians to a vital tool in economic planning, environmental debates, ecological modeling, and indeed almost every science.As computer simulation has grown in importance, philosophers have felt the need to address what can and cannot be learned from simulations.Some claim simulation has nothing empirical to tell us.More claim an entirely new epistemology is required to make sense of simulation science, that simulation science should well and truly reconstruct how we make sense of science.

Biography:

Alan Dorin enjoys exploring science through art, and art through science as well as each of these disciplines for its own sake. His interests included at one stage or another: Applied Mathematics;Animation & Interactive Media; Biology (artificial/synthetic and natural); Computer Science; History (especially of Science and Art);Mechanics; Music; Philosophy; Self-assembly; Visual Art and the links that bind all fields together. Alan occasionally finds time to do his own research, make animations, computer music, write papers, give lectures, organise conferences, supervise students - the usual things people working in academia do and as many as possible they don't. Alan co-directs the Centre for Electronic Media Art in the Faculty of Information Technology at Monash University, Australia. He spends his non-work time racing bicycles and scaling mountains.

Abstract:

Virtual ecosystems are agent-based software simulations of the interactions between organisms and an abiotic environment. They range in scope from highly abstract systems based on cellular automata-like transition rules, to literal models of specific organisms and their habitats. In each case the typical approach to their construction and application runs roughly as follows: construct a data-structure to represent the agents and another to represent the environment;specify the possible inter-agent interactions; specify the agent- environment interactions; specify the initial conditions... and then let the simulation run whilst observing the results. The aim is to demonstrate that the system is capable of exhibiting some form of "emergent" behaviour that supervenes upon the initial conditions and rules of interaction that the programmer has established. The system may be used to test hypotheses about the necessary and sufficient conditions for the emergence of such phenomena in the natural world. This seminar presents a (biased) overview of approaches adopted by researchers who have explored virtual ecosystems. It demonstrates a few diverse systems that the presenter has himself developed along with his personal views as to why these models are "interesting". A preview of a new virtual ecosystem being developed by the author is also presented for comment.

Biography:

Maria Garcia de la Banda received in 1994 the Best PhD Award from the Faculty of Computer Science at the Technical University of Madrid, Spain. In 1997 she was awarded a Logan Fellowships at Monash University, where she is now an associate professor at the Faculty of Information Technology. While her research is mainly focused on the design, compilation and applications of constraint programming languages, she is also interested in bioinformatics.

Abstract:

This talk discusses a dynamic programming solution to the problem of minimizing the maximum number of open stacks, a resource-planning problem that appears not only in manufacturing industries but also in other interesting situations such as scheduling the execution of SAT clauses. Starting from a call based dynamic program, we show a number of ways to improve the dynamic programming search, preprocess the problem to simplify it, and to determine lower and upper bounds. We then explore a number of search strategies for reducing the search space. The final dynamic programming solution is, highly effective.

Biography:

http://www.cs.mu.oz.au/~pjs/

Abstract:

The G12 project at National ICT Australia (NICTA)is an ambitious project to develop a software platform for solving large scale industrial combinatorial optimisation problems. The core design involves three languages: Zinc, Cadmium and Mercury (Group 12 of the periodic table). Zinc is a declarative modelling language for expressing problems, independent of any solving methodology. Cadmium is a mapping language for mapping Zinc models to underlying solvers and/or search strategies, including hybrid approaches. Finally, existing Mercury will be extended as a language for building extensible and hybridizable solvers. The same Zinc model, used with different Cadmium mappings, will allow us to experiment with different complete, local, or hybrid search approaches for the same problem. This talk will explain the G12 global design, the G12 objectives, and our progress so far.

Biography:

Edwin Hancock has been with the Department of Computer Science at the University of York since July 1991 where he is currently Professor of Computer Vision. His research interests are in the areas of computer vision and pattern recognition and he has externally funded projects in the areas of terrain analysis, sensor fusion and visual learning. Most of what Edwin does involves Bayesian models or optimisation. An important thread running through his research is how to use relational graphs to represent visual information and how to match inexact graphs. Ongoing research topics include machine learning in computer vision, graph-spectral methods, 3D surface reconstruction from 2D images and quantum computing. Prior to becoming active in the field of computer vision and pattern recognition, Edwin did a PhD and a postdoc in high-energy nuclear physics, which involved experiments at CERN and Stanford. His first degree was in theoretical physics. Professor Hancock is an Associate Editor of the journals IEEE Transactions on Pattern Analysis and Machine Intelligence and Pattern Recognition. He has also been a Guest Editor for Image and Vision Computing and in 1994 chaired the 1994 British Machine Vision Conference. In 1995, he co-founded the EMMCVPR Workshop series with Marcello Pelilllo and has served in chair and co-chair positions on numerous other conference committees.

Edwin has published widely in the field of computer vision and pattern recognition. He has had approximately 90 journal papers and 350 refereed conference papers published and received the best paper medal and an outstanding paper award for contributions to the journal Pattern Recognition. He is a member of UKCRC, a fellow of the IAPR and Governing Board Member of the IAPR.

Abstract:

This talk will focus on two different approaches to learning generativemodels of graph-structure. The first of these is information theoretic and demonstrates how description length criteria can be used to learntree-structure. Here we will show how to code tree structure and how thedescription length criterion can be used to cluster trees by fitting a mixture of tree-unions to a set of sample data. The second approach is a geometric one and utilises concepts from spectral graph theory. We makeuse of the Laplacian spectrum to embed the nodes of a graph in a vector,and show how the geometry of this embedding can be used to construct a generative model for graph structure. The talk will be illustrated by results for object recognition.

Biography:

http://www.csse.monash.edu.au/~lloyd/

Abstract:

Compression can be rather addictive. It has an obvious ``figure ofmerit'' -- the compressed size of some standard data-set, and it is natural to strive to beat your competitors on this measure. Biological sequences are hard to compress; the more compression the better butthere is usually limited value in fighting over the third significantdigit. Other properties of a compression method can be just as importantor more important. This talk gives some reasons why it is challenging,interesting and useful to compress biological sequences. It alsopresents two simple models for compressing biological sequences (apossible sub-addiction in compression is to complicated models, butsimple is often good); we get good results for DNA and protein.

Biography:

Sarah Boyd graduated in 2000 with a double major in computer science and biochemistry, and an honours degree in computer science (Monash University).In 2005 she obtained her PhD, focusing on computational methods for predicting protease specificity (also at Monash University).In 2005-2006, she worked as a research fellow working on more computational methods for understanding protease specificity (still at Monash University), and at the end of 2006 was awarded an ARC Discovery Grant and Australian Postdoctoral Fellowship, through which she is now employed to do further research, at Monash University, on computational methods for protease research. Sarah is interested in proteases and bioinformatics, and thinks Monash is an ok place to work.

Abstract:

Proteases are enzymes that modify proteins. They are ubiquitous in all forms of life, controlling biological process as diverse as growth and development, digestion, immunity, and ultimately cell death. In addition, they are implicated in many diseases including autoimmune diseases, cancers, infections and blood disorders. Thus, in-depth understanding how proteases function is essential for us to learn how biological processes are regulated, how diseases develop, and how we can design effective protease inhibitors to treat such diseases. Unfortunately, protease research is time-consuming and costly, because of the complexity of the proteases themselves, and also the processes that they regulate. To address these problems, we have developed a computational system called PoPS: Prediction of Protease Specificity (http://pops.csse.monash.edu.au/) allows researchers to investigate and understand how their favourite protease functions. In particular, PoPS allows users to create a mathematical model of any protease based on any source of knowledge from laboratory experiments through to 'expert knowledge'. The PoPS model can then be used with the PoPS tools to reason about known functions of the protease, and/or predict novel roles. In this talk, I will describe the PoPS protease model and computational tools, with examples of how this system has been applied to some specific medical research projects. I will then describe new computational challenges that we are facing in the analysis and interpretation of experimental data, and understanding and prediction of complex biological processes.

Monash Contact: Maria Garcia de la Banda

Biography:

Sven Koenig is an associate professor in computer science at the University of SouthernCalifornia. Most of his research centres around techniques for decision making (planning and learning) that enable single situated agents (such as robots or decision-support systems) and teams of agents to act intelligently in their environments and exhibit goal-directed behaviour in real-time, even if they have only incomplete knowledge of their environment, imperfect abilities to manipulate it, limited or noisy perception or insufficient reasoning speed. He is the recipient of an NSF CAREER award, an IBM Faculty Partnership Award, a Charles Lee Powell Foundation Award, a Raytheon Faculty Fellowship Award, and an ACM Recognition of Service Award, among others. He co-founded Robotics: Science and Systems (a highly selective robotics conference), was conference co-chair of the 2002 Symposium on Abstraction, Reformulation, and Approximation, conference co-chair of the 2004 International Conference on Automated Planning and Scheduling, program co-chair of the 2005 International Joint Conference on Autonomous Agents and Multi-Agent Systems and program co-chair of the 2007 AAAI Nectar program. Additional information about Sven can be found on his webpages: idm-lab.org.

Abstract:

In this talk, I will give an overview of our research on market mechanisms for the allocation of resources in cooperative domains. As example, I will use exploration tasks where a team of mobile robots needs to visit a number of given targets in known or partially unknown terrain. An important characteristic of these multi-robot routing tasks is that the assignment of targets to robots can turn out to be suboptimal as the robots gain more information about the terrain. Auctions promise to assign and re-assign targets to robots efficiently in terms of both the required amount of computation and communication since information is compressed into numeric bids that the robots can compute in parallel. I will discuss the advantages and disadvantages of different auction mechanisms, including recent theoretical results that show that sequential single-item auctions can provide constant factor performance guarantees in known terrain even though they run in polynomial time. Time permitting, I will also discuss generalizations of sequential single-item auctions, such as sequential single-item auctions with bundles. This is joint work with D. Kempe, P. Keskinocak, M. Lagoudakis, V. Markakis, A. Meyerson, C. Tovey and our students.

Monash Contact: Ann Nicholson

Biography:

Falk Schreiber graduated (1997) and obtained a PhD in Computer Science(2001) from the University of Passau (Germany). After this he worked as a Research Fellow and Lecturer at the University of Sydney (Australia). Since 2003 he is head of a research group at the Leibniz Institute of Plant Genetics and Crop Plant Research Gatersleben (Germany). His research is focused on the areas of algorithms, visualization, and bioinformatics; particularly in application-oriented and interdisciplinary directions.

Abstract:

Systems biology is a new field in biology that aims at the understanding of complex biological systems, such as a complete cell. It has emerged in the light of the availability of modern high-throughput technologies,which result in huge amounts of molecu lar data regarding life processes. This data is often related to, or even structured in, the form of biological networks such as metabolic, protein interaction and gene regulatory networks. Network-related data analysis and exploration methods help scientists to extract information out of the data and thus are very useful for building sophisticated research tools. This presentation gives a brief introduction into molecular biological processes and systems biology. It discusses examples of the analysis of fundamental biological networks and their user-friendly visualisation. These examples range widely from analysing experimental data in the context of the underlying biological networks to structural analysis and subsequent visualisation of biological networks based on motifs, clustering and centralities. Finally we consider new directions and questions in analysing and visualising these large and complex networks.

Monash contact: Kim Marriott