seminars
RMITOpt Talk: Extending de Bruijn’s Identity – Prof. Bill Moran
Talk by Dr. Walter Kozlowski in April
![]() |
Speaker: Dr. Walter Kozlowski, Network Evolution 2020 Infrastructure,Telstra Title: Common fixed points of monotone Lipschitzian semigroups in Banach spaces Date and Time: Friday, April 13th, 3.00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus (To connect via visimeet please contact rmitopt@rmit.edu.au) |
Abstract: We will discuss the existence of common fixed points of monotone Lipschitzian semigroups in Banach spaces under the natural condition that the images under the action of the semigroup at certain point are comparable to the point. In particular, we prove that if one map in the semigroup is a monotone contraction mapping, then such common fixed point exists. In the case of a monotone nonexpansive semigroup we prove the existence of common fixed points if the Banach space is uniformly convex in every direction. We explain why these assumptions are substantially weaker than those made in the pre-existing literature. We will also touch upon some non-trivial applications, as well as on recent research directions in the field.
Bio: Dr Walter Kozlowski – publishing in Mathematics as Wojciech M. Kozlowski – has been actively involved in research activities in Mathematics and Applications since 1980s with a particular interest in functional analysis, fixed point theory and applications. A Fulbright scholar, an author of a monographic book “Modular Function Spaces”, co-author of recently published book “Fixed Point Theory in Modular Function Spaces” and of numerous scientific papers, he has held several academic posts at the universities around the world, currently as an Adjunct Professor at the School of Mathematics and Statistics of the University of New South Wales and a research collaborator of CARMA (Research Centre for Computer-Assisted Research Mathematics and its Applications) at the University of Newcastle, NSW, Australia. In parallel, Walter has been pursuing a professional career in Information and Communication Technology with over 25 years of increasingly responsible experience within leading global organisations including IBM, Telstra, and Hewlett-Packard. In this capacity, he has been leading large-scale digital transformation projects for clients in the telecommunications and media industries worldwide, as well he has published several industry related business and technology articles. Walter has been awarded by The Open Group the title of the Distinguished IT Architect in the category of Chief Architect. Walter also participates in the PhD Industry Mentoring programme organised by the La Trobe University in Melbourne. In 2017 Walter has assumed a position of the Chief Architect of the Telstra Network Function Virtualisation program, in which role he leads the technology transformation of towards the software controlled, virtualised network.
|
RMITOpt Seminar this Friday the 23rd March at 1.30pm – Dr. Anand Deopurkar, ANU
![]() |
Speaker: Dr. Anand Deopurkar
Title: The algebra and geometry of algebraic varieties Date and Time: Friday, March 16th, 1:30pm – 2.30pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus (To connect via visimeet please contact rmitopt@rmit.edu.au) |
Abstract: Abstract: An algebraic variety is the set of common zeros of a collection of polynomial equations. A central theme in algebraic geometry is to understand the connection between the geometric properties of the zero set (for example, curvature) and the algebraic properties of the polynomials (for example, degrees). I will explain known and conjectural results in this area with lots of examples.
Bio: Anand is lecturer at the Australian National University. Before coming to ANU, he was a post-doc at Columbia University and the University of Georgia, and before that, a graduate student at Harvard. Anand studies moduli spaces of algebraic varieties. That is, he tries to understand how algebraic varieties deform when the defining equations are perturbed. In addition to algebraic geometry, he enjoys thinking about problems in algebra, number theory, and representation theory. |
Talk by Bertrand Jouve, March the 16th
![]() |
Speaker: Bertrand Jouve
Title: Using graph theory for the study of medieval social network Date and Time: Friday, March 16th, 3:00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus (To connect via visimeet please contact rmitopt@rmit.edu.au) |
Abstract: We show how classical and new tools of graph theory may be used for studying medieval sources. The large-scale processing of information contained in agrarian contracts between 1250 and 1550 allows to propose forms of social organizations of the peasant world of the south of France that bring new points of view to some major questions posed by historians.
Bio: Bertrand Jouve is a senior researcher at the French National Centre for Scientific Research (CNRS). He studied pure mathematics, passed the agrégation in Mathematics and, in 1999, was awarded a doctorate (PhD) in Cognitive Sciences under the supervision of Pierre Rosenstiehl (mathematician) and Michel Imbert (neuroscientist) at the Higher School of Social Sciences (EHESS, Paris). First an Assistant Professor in Toulouse, then a Full-Professor of Applied Mathematics in Lyon, he was appointed Senior Researcher by the CNRS 53 interdisciplinary Commission in 2014. Bertrand Jouve’s work focuses on the construction of mathematical models of graph theory to be used for the analysis of large-scale interaction networks. Over a decade, he has worked at the boundary of neurosciences on the topological structure of cortico-cortical neural connections of monkey brains. Since 2004, he has been working at the boundary of Social Sciences and Humanities. He first collaborated with linguists on the structure of language dictionaries, with sociologists to analyze large-scale social networks and with historians with a view to reconstructing former complex social networks. Bertrand Jouve created the Toulouse Institute for Complex System Studies (XSYS) in 2014. Bertrand Jouve is engaged in the national administration and strategy of French research. He was a member of the CNRS board between 2011 and 2016 and member of the CNRS Committee for Intensive Computing. Since June 2016, he has been President of the National Network of 23 Houses of Social Science and Humanities (RNMSH) which is a part of the French national strategy for excellence in Social Sciences and Humanities. |
Talk by Asen Dontchev in early March
![]() Speaker: Asen Dontchev Title: Constrained Spline Interpolation Date and Time: Friday, March 2nd, 3:00pm – 4.00pm Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus (To connect via visimeet please contact rmitopt@rmit.edu.au) |
Abstract: It is well known that the interpolating cubic spline can be viewed as a solution of an optimal control problem. This fascinating connection between the classical approximation theory and the theory of optimal control has been explored by researches from both sides in different times and often independently. In this talk I will review some developments on spline interpolation with constraints. In particular, I will show how a long-standing open question about super-linear convergence of Newton’s method for best convex interpolation has been settled.
Final Presentation Healthier/RMIT by Iman Roozbeh at RMIT January the 16th
![]() |
Speaker: Iman Roozbeh
Title: Improving the optimisation of automated meal plan generation through a combination of specific health information and personal preferences Date and Time: Tuesday, January 16, 10:30am – 11:30am Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus |
Abstract: Proper dietary guidance is a critical component of chronic disease management. However, just 5,000 dietitians operate in Australia to serve the estimated 7 million people currently living with chronic disease. Our current approach to delivering health care is not coping now and it certainly won’t in the future. Dietitians face the challenges of integrating complex scientific information, lifestyle and food preferences, applied to a range of medical conditions. Dietitians cannot provide personalised meal plans to every client because the process is too time consuming and many do not have adequate training.
To address these issues, Health Delivered is developing a dietary management platform to tackle the growing rates of obesity, diabetes and other chronic health issues by building customised meal plans for patients. The platform has been created to build meal plans that account for social, personal and medical history as well as health goals and personal preferences, in an efficient manner and assist dieticians to see more number of clients and focus on quality of client care.
Katharine Turner, Persistent homology rank function
Speaker: Dr Katharine Turner
Australian National University
Title: Persistent homology rank function
Date and time: Monday 13 November 2017, 10:30am
Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus
Abstract:
Persistent homology is a tool for capturing how topological features evolve over an increasing family of spaces. Commonly these spaces are taken to be the unions of balls of increasing radii about some finite set of points. Using this scaling parameter we can summarise geometric information as a topological summary statistic. In this talk I will introduce persistent homology and define the persistent homology rank function which, as a functional summary lying in a Hilbert space, enables us to perform statistical analysis such as principal component analysis. I will present some applications including testing complete spatial randomness of spatial point patterns, and comparing experimental sphere packings and colloid data under different temperatures. This talk is based on work with Vanessa Robins.
CIAO Talk, Reinier Diaz Millan: On the splitting optimization problem with enlargement

Speaker: Dr Reinier Diaz Millan
Federal Institute of Goias
Title: On the splitting optimization problem with enlargement
Date and time: 11:30am, Thursday 16 November 2017
Location: Room T121, Mt Helen Campus (this is at Federation University, RMIT crowd please connect via Visimeet).
Abstract:
In this paper, we present two approximate versions of the forward-backward splitting method for solving the minimization problem. In both cases, the objective function is the sum of two convex functions, maybe not differentiable. The algorithms involve, at each iteration, inexact evaluations of the backward operator and approximate subgradients of the functions (namely: the ε-subgradients). The first method considers an absolutely summable error criterion, whereas the second method uses a relative error criterion recently introduced for approximating proximal operators. Various stepsize rules are considered, including both diminishing and non-vanishing stepsizes, and convergence in objective values and convergence to a neighbourhood of the optimal set are obtained. The convergence analysis of the two methods shares underlying elements. Read more
Kate Smith-Miles: Optimization in the Darkness of Uncertainty: when you don’t know what you don’t know, and what you do know isn’t much!

Speaker: Prof. Kate Smith-Miles
University of Melbourne
Title: Optimization in the Darkness of Uncertainty: when you don’t know what you don’t know, and what you do know isn’t much!
Date and time: Monday 6 November 2017, 3:30–4:30pm
Location: Building 8 Level 9 Room 66 (AGR) RMIT City campus
Abstract:
How do we find the optimal solution for a constrained multiobjective optimisation problem when we have no analytical expression for the objective functions, and very limited function evaluations within the huge search space due to the expense of measuring the objective functions? Calculus can’t help you, and trial and error is not an option! This talk will describe a common practical optimisation problem found in many industrial settings with these challenges, and introduce some methods for expensive black-box optimisation. Finally, we will address the question of how best to evaluate the performance of such methods by generating new test instances with controllable characteristics. Read more