 |
Speaker: Dr. Matthew Tam, the School of Mathematics and Statistics at the University of Melbourne.
Title: Algorithms derived from dynamical systems Date and Time: Friday, March 13th, 3.00pm – 4.00pm, 2020 (Talk & Q/A) Location: AGR Building 15, level 03, room 10 (Request for remote Zoom connect andy.eberhard@rmit.edu.au) |
| Abstract: The study of continuous time dynamical systems associated with iterative algorithms for solving optimisation problems has a long history which can be traced back at least to 1950s. The relationship between the continuous and discrete versions of an algorithm provides a unifying perspective which gives insights into their behaviour and properties. In this talk, I will report on new algorithms for solving minmax problems which were discovered by exploiting this connection.
Bio: Matthew Tam is Lecturer in Operations Research and a DECRA Fellow in the School of Mathematics and Statistics at the University of Melbourne. He received a PhD from the University of Newcastle in 2015 under the supervision of Jonathan Borwein, where he worked on iterative projection algorithms for optimisation. He then moved to the University of Göttingen (Germany) where he was a post-doctoral researcher with Russell Luke, supported initially by DFG-RTG2088 (“Discovering structure in complex data”) and later by a fellowship from the Alexander von Humboldt Foundation. Prior to joining the University of Melbourne, he was Junior Professor for Mathematical Optimisation within the Institute for Numerical and Applied Mathematics, also at the University of Göttingen. |
|
news
RMITOpt Seminar: Dr. Patrick Johnstone, MSIS Department of the Rutgers Business School
 |
Speaker: Dr. Patrick Johnstone, MSIS Department of the Rutgers Business School.
Title: Projective Splitting: A New Breed of First-Order Proximal Algorithms Date and Time: Friday, February 28th, 3.30pm – 4.30pm, 2020 (Talk & Q/A) Location: AGR Building 15, level 3, room 10 (Request for remote connect andy.eberhard@rmit.edu.au) |
| Abstract: Projective splitting is a proximal operator splitting framework for solving convex optimization problems and monotone inclusions. Unlike many operator splitting methods, projective splitting is not based on a fixed-point iteration. Instead, at each iteration a separating hyperplane is constructed between the current point and the primal-dual solution set. This gives more freedom in terms of stepsize selection, incremental updates, and asynchronous parallel computation. Despite these advantages, projective splitting had two important drawbacks which we have rectified in this work. First, the method uses calculations entirely based on the proximal operator of the functions in the objective. However, for many functions this is intractable. We develop new calculations based on forward steps – explicit evaluations of the gradient – whenever the gradient is Lipschitz continuous. This extends the scope of the method to a much wider class of problems. Second, no convergence rates were previously known for the method. We derive an O(1/k) rate for convex optimization problems, which is unimprovable for this algorithm and problem class. Furthermore, we derive a linear convergence rate under certain strong convexity and smoothness conditions.
Bio: Patrick R. Johnstone is a postdoctoral associate in the MSIS Department of the Rutgers Business School where he is advised by Prof. Jonathan Eckstein. In May 2017 he received his PhD in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign (advised by Prof. Pierre Moulin). He also received the MSc degree in ECE from UIUC. He received the BSc degree in Electrical Engineering from the University of New South Wales (UNSW) in Sydney, Australia. At UNSW he received the university medal, given to the top student in electrical engineering. He has worked as a research intern at Qualcomm Research, Rambus Labs, and CSIRO. For his work at Qualcomm, he received the Roberto Padovani award for outstanding interns. His research interests are in continuous optimization, first-order proximal splitting methods, parallel and distributed algorithms, machine learning, and signal processing. |
|
RMIT Optimisation Seminar: Dr. Inmaculada Flores Garcia., Complutense Unversity of Madrid
|
Speaker: Dr. Inmaculada Flores Garcia., Complutense Unversity of Madrid
Title: Multi-criteria optimization for disaster evacuation of people and relief distribution Date and Time: Friday, November 1 st, 3.00pm – 4.00pm Location: Megaflex (near RMIT Connect), Building 8 Level 4 Room 13, RMIT City campus |
| Abstract: Disasters and their consequences strike the world population from the beginning of history and their occurrences have an upward trend. Palliating the effects of a disaster is the main objective of the work that will be presented, which is focused on the supported evacuation of the population from the affected area to safer places, with the aim of preventing people from suffering the consequences of the catastrophic event. Furthermore, and once people are located at a safe place, basic supplies must be delivered in order to assure that the evacuated population have their basic necessities covered after the occurrence of a disaster. To tackle this complex problem, a mathematical mixed integer programming model based on time dependence is presented. The problem is multimodal in terms of different fleets of vehicles, types of affected population and of commodities to be distributed.
Bio: Inmaculada has a 5-year degree on Mathematics by the University of Granada, including a year course at the Technische Universität München, and a master degree on Mathematical Engineering at the Complutense University of Madrid. After 6 years working on Data Analysis and Information Security in two important companies at Spain, she became a PhD student and mid-time professor at the Faculty of Mathematical Sciences and at the Faculty of Commerce and Tourism in the Complutense University of Madrid. Currently, she belongs to the Statistics and Operations Research department (Fac. Mathematical Sciences, UCM) as Project Research Staff. |
|
OPTIMISATION METHODS IN WILDFIRE EMERGENCY MANAGEMENT
We are happy to invite contribution and participation to the Workshop
OPTIMISATION METHODS IN WILDFIRE EMERGENCY MANAGEMENT
November 11 – RMIT University, city campus
co-funded by GEO-SAFE , AMSI and AustMS/ANZIAM
The problems arising in emergency management have to take into account a high level of uncertainty. In wildfire management, in particular, weather conditions, the stress level of vegetation, location of ignition and fire behaviour dynamics constitute the main sources of uncertainty. In order to protect communities in a safer and more efficient way, emergency management can use optimisation approaches that contribute to reduce uncertainty. Depending on the type of uncertainty and the objectives, the resulting optimisation problems can be classified in different paradigms including stochastic optimisation, robust optimisation, probabilistic combinatorial optimisation or online optimisation.
The workshop will screen the state of research in optimisation for wildefire emergency management and by extension, of any mathematical modelling approach aiming to improve the efficiency of emergency management. Details and updates are available on the workshop’s website.
Two invited international experts on wildfire emergency management and optimisation methodologies will give a lecture:
Prof. Cristina Vega, Universitat de Lleida, Spain: “Towards a landscape fire management strategy in Southern Europe based on risk analysis: Unresolved spatial and mathematical issues”
Assoc. Prof. Aurélie Thiele, Southern Methodist University, Dallas, USA: “An introduction to robust optimization with applications to emergency management”
Who can participate? The workshop is primary dedicated to PhD students and early career researchers. But everybody interested in the topic is welcome.
Location: Melbourne, Vic., RMIT University – Swanston Academic Building 445 Swanston Street – Level 5, room 12.
How to participate? Is you want to attend the workshop, send an email with your details to M. Demange (marc.demange@rmit.edu.au). Participation is free but registration is required (deadline 4th of November).
If you want to present an abstract, send it to Marc Demange (marc.demange@rmit.edu.au) before October 18th, 2019 addressing the following items:
- Real problem that you are addressing.
- Methodological approach that you are using to solve that problem.
- Main challenges identified from the approach you are using.
- Final outcome expected from your research and how this will be used to solve real life problems.
OPPORTUNITIES FOR PARTICIPANTS
Australian attendees from AMSI Member organisations have access to funding via the AMSI Travel Fund. Students or early career researchers from AMSI member universities without access to a suitable research grant or other source of funding may apply to the Head of Mathematical Sciences for subsidy of travel and accommodation out of the departmental travel allowance. We recommend applying as soon as possible. http://research.amsi.org.au/travel-funding/
The Workshop will be followed by GEO-SAFE International Wildefire Conference from November 12 to November 15. Early bird rates available until October 18.
Organisers: Marc Demange, John Hearne, David Ellison (RMIT, Australia), Marta Yebra (Australia National University, Australia), Jagannath Aryal (University of Tasmania, Australia), Núria Prat-Guitart (Pau Costa Foundation, Spain).
RMIT Optimisation Seminar: Dr. Alysson M. Costa, University of Melbourne
 |
Speaker: Dr. Alysson M. Costa, University of Melbourne
Title: Load consolidation for freight distribution in urban environments Date and Time: Friday, October 4th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
| Abstract: I will present three decision problems in freight systems. All problems focus on load consolidation aiming to improve distribution efficiency in urban environments. First, I will consider the use of a single cross-dock from an operational perspective, focusing on the decision of when to dispatch outbound trucks. The second problem also includes cross-docks but now studies the problem from a network design perspective, aiming to decide on cheapest distribution routes. Finally, I will present a modelling approach for the design of a collaborative distribution network.
This talk presents the results obtained by Pedro Castellucci in his doctoral thesis (University of Melbourne / University of Sao Paulo). This work also has the collaboration of Associate Professors Russell Thompson (University of Melbourne) and Franklina Toledo (University of Sao Paulo). Bio: Alysson M. Costa is a Senior Lecturer in Operations Research at the School of Mathematics and Statistics – University of Melbourne. He is interested in theory and applications of Optimisation. Throughout his career, he has worked extensively with mixed integer programming (modelling and solution methods) applied to problems in different areas such as environmental water management, disaster relief operations, educational timetabling, crop rotation, assembly line balancing and city logistics, among others. He received his PhD in 2006 from HEC Montreal / University of Montreal – Canada. His thesis, titled “Models and algorithms for two network design problems”, received the Cecil Graham doctoral dissertation award from the Canadian Applied and Industrial Mathematics Society. Before that, he received MSc. and B.Eng. Degrees in Electrical Engineering from the State University of Campinas – Brazil. |
|
RMIT Optimisation Seminar: Dr. Rhys Bowden, University of Melbourne
 |
Speaker: Dr. Rhys Bowden, University of Melbourne
Title: Consensus versus scaling for cryptocurrency blockchains Date and Time: Friday, September 20th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
| Abstract: Bitcoin and other cryptocurrencies maintain a distributed global ledger in the form of a blockchain. Blocks are like pages in the ledger; each block is linked to the most recent prior block in the chain. Blocks are generated at random times at random mining computers in the Bitcoin network, then propagated to the other miners and the remainder of the network.
In order to increase the rate at which transactions are included in the blockchain, some proposals suggest increasing the rate at which blocks are mined relative to how long it takes them to propagate through the network. This can result in different versions of the blockchain being present at different miners. Miners can only mine on top of blocks they know about, so instead of a linear chain of blocks there’s now a tree. We model this situation to find out how this “blocktree” grows and what that means for global agreement on the contents of the ledger. Bio: Dr Rhys Bowden obtained a PhD in mathematics from the University of Adelaide in 2015 for a thesis titled “Link Loss Tomography and Topology Synthesis”. Since then, he has worked at the University of Melbourne, first in the Department of Electrical and Electronic Engineering and then in the School of Mathematics and Statistics. He is interested in modelling stochastic processes (particularly on graphs), computer network measurement, network tomography, cryptocurrencies and blockchains. |
|
RMITOpt Seminar: Dr. Gregorio Tirado Domínguez – Universidad Complutense de Madrid
 |
Speaker: Dr. Gregorio Tirado Domínguez – Universidad Complutense de Madrid
Title: Multi-criteria Models and Heuristics for the Post-disaster Distribution of Humanitarian Aid Date and Time: Friday, August 16th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
| Abstract: The distribution of humanitarian aid is one of the most important logistic operations to be carried out to assist the affected population after the strike of a disaster. In this context, it is crucial to minimize the response time, as well as taking into account operation costs and designing reliable and safe itineraries that lead to equitable distribution plans. This situation can be modelled in a natural way by using multi-criteria mathematical programming models. In particular, this seminar will focus on the development of a lexicographical goal programming model for the distribution of humanitarian aid, together with the design of appropriate solution methods. Some results regarding their application to several realistic case studies will be presented to illustrate their performance.
Bio: Gregorio Tirado Domínguez is an Associate Professor at the Departament of Actuarial and Financial Economics and Statistics of the Faculty of Economics and Management of the University Complutense of Madrid. He holds a PhD degree in Mathematics from the University Complutense of Madrid since 2009. His main research lines comprise combinatorial optimization, logistics (vehicle routing problems and humanitarian logistics, in particular), mathematical programming and heuristics. |
|
RMITOpt Seminar: Dr. Nadia Sukhorukova, Swinburn University
 |
Speaker: Dr. Nadia Sukhorukova, Swinburn University
Title: Chebyshev approximation problems and Remez method applicability. Date and Time: Friday, August 2nd, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
| Abstract: In this presentation, I am going to talk about three types of Chebyshev approximation problem: approximation by univariate polynomials, multivariate polynomials and rational functions. The objective function in the first two problems are convex and the corresponding optimisation problems can be solved by applying a general convex approximation method, while the third type is a quasi-convex and needs a specially developed method to solve it. In this talk, I will demonstrate the transformation of the problem and highlight surprising similarities passing from one formulation to another.
Bio: Dr. Sukhorukova is holding two PhD degrees: one from the University of Ballarat (currently, Federation University Australia), completed in 2004 and another one is from Saint-Petersburg State University (Russia) completed in 2006. Dr. Nadia Sukhorukova is working in the area of Mathematical Optimisation and its application to data analysis, data mining, signal processing, location-allocation problems and other real-life applications. Dr. Sukhorukova is also working on a number of mathematical problems appearing in the area of Chebyshev (uniform) approximation (ARC Discovery grant DP180100602). |
|
RMITOpt Seminar: Dr. Chayne Planiden, University of Wollongong
 |
Speaker: Dr. Chayne Planiden, University of Wollongong
Title: A Derivative-free VU-algorithm for Convex Minimisation Date and Time: Friday, July 19th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
| Abstract: The VU-algorithm is a superlinearly convergent method for minimizing nonsmooth functions. The algorithm separates the space into the V-space and the orthogonal U-space, such that the nonsmoothness of the objective function is concentrated on the V-space, and on the U-space the function behaves smoothly. This structure allows for an alternation between a fast, Newton-like step in the U-space and a proximal-point step in the V-space, thus increasing convergence speed.
We establish a derivative-free variant of the VU-algorithm, the first of its kind, and show global convergence. We provide numerical results from a proof-of-concept implementation, which demonstrate the feasibility and practical value of the approach. Bio: Chayne Planiden received his Ph.D. from University of British Columbia in Kelowna, Canada last year. He works at University of Wollongong and specialises in nonsmooth optimisation, including the Moreau envelope and proximal mapping, proximal point algorithms, and derivative-free optimisation methods. |
|
RMITOpt Seminar: Bill Moran, Melbourne University
 |
Speaker: Prof. Bill Moran, Melbourne University
Title: Random Thoughts on Randomization Date and Time: Friday, May 24th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus (To connect via Zoom please contact andy.eberhard@rmit.edu.au) |
| Abstract: Surprisingly, the randomization of a problem space can often make it easier to solve than the deterministic original. This trick occurs across a range of areas of mathematics, statistics, computer science, and even engineering. Early instances are Borel’s Theorem on normal numbers and mixed strategies in game theory. Other examples include Erdos-Renyi graphs and their work on additivity properties of sequences of integers. I will give a survey of the various ideas and, as time permits, discuss a range of applications to: number theory, graph theory, signal processing, optimization, computer science, game theory, and statistics.
Bio: Professor Bill Moran (M’95) currently serves, since 2017, as Professor of Defence Technology in the University of Melbourne. From 2014 to 2017, he was Director of the Signal Processing and Sensor Control Group in the School of Engineering at RMIT University, from 2001 to 2014, a Professor in the Department of Electrical Engineering, University of Melbourne, Director of Defence Science Institute in University of Melbourne (2011-14), Professor of Mathematics (1976–1991), Head of the Department of Pure Mathematics (1977–79, 1984–86), Dean of Mathematical and Computer Sciences (1981, 1982, 1989) at the University of Adelaide, and Head of the Mathematics Discipline at the Flinders University of South Australia (1991–95). He was Head of the Medical Signal Processing Program (1995–99) in the Cooperative Research Centre for Sensor Signal and information Processing. He was a member of the Australian Research Council College of Experts from 2007 to 2009. He was elected to the Fellowship of the Australian Academy of Science in 1984. He holds a Ph.D. in Pure Mathematics from the University of Sheffield, UK (1968), and a First Class Honours B.Sc. in Mathematics from the University of Birmingham (1965). He has been a Principal Investigator on numerous research grants and contracts, in areas spanning pure mathematics to radar development, from both Australian and US Research Funding Agencies, including DARPA, AFOSR, AFRL, Australian Research Council (ARC), Australian Department of Education, Science and Training, and Defence Science and Technology, Australia. His main areas of research interest are in signal processing both theoretically and in applications to radar, waveform design and radar theory, sensor networks, and sensor management. He also works in various areas of mathematics including harmonic analysis, representation theory, and number theory. |
|