Welcome (Carolyn Robinson, Business Manager)
& Acknowledgement of Country (Kelsey Moore, Senior Project and Communications Officer)
Much of the research that is generated by Universities and Research Centres is driven by PhD students. In this presentation I will showcase some of the work done by some of my recent PhD students as part of their thesis. These vignettes will focus on research in “symbolic data analysis”, which aims to perform statistical analyses for data that have been summarised into distributional form, such as random rectangles, random histograms, and other distributions. I will illustrate the ideas behind each project through both simulation studies and real-world environmental applications.
Speaker: Scott Sisson
Scott is the Director of the UNSW Data Science Hub (uDASH) and Professor in Statistics at UNSW.
uDASH is a major strategic initiative of UNSW Science, which cultivates and promotes foundational and applied research in Data Science across a range of issues and industries. The Hub provides a world-class environment, with access to state-of-the-art data visualisation and computing facilities and enables the creation, development and deployment of transformative data-driven decision-making tools that help address current and future societal challenges.
As Professor of Statistics at UNSW, Scott is internationally recognised for his work in computational and Bayesian statistics, particularly for developing inferential techniques for computationally intractable models and challenging data. He has been awarded the P.A.P. Moran Medal (Australian Academy of Science), the G. N. Alexander Medal (Engineers Australia) and the J. G. Russell Award (Australian Academy of Science).
Scott is a former Deputy Director of the ARC Centre of Excellence for Mathematical and Statistical Frontiers, and ARC Future Fellow. Recently he was President of the Statistical Society of Australia, and Head of Statistics at UNSW.
Dr Ranran (Monica) Bian received her PhD from The University of Auckland and is currently working as a Research Fellow at ARC Training Centre in Data Analytics for Resources and Environments (DARE) of The University of Sydney. Prior to joining The University of Sydney, she worked as a postdoctoral researcher at The University of Adelaide.
Dr Bian’s PhD focused on heterogeneous network mining and analysis, where novel algorithms for community discovery, ranking, dynamic embedding and change modelling in different heterogeneous networks were developed. The key areas of her research interests include representation learning for large-scale dynamic heterogeneous networks, social network analysis, applying and adapting network embedding and analysis techniques in resource and environment management.
My research interest is in quantifying and communicating uncertainty within environmental modelling. I am particularly interested in the application of probabilistic machine learning techniques to the polar Earth and Climate Sciences: Currently I am working on Bayesian Optimisation for subsurface inversion of Antarctica’s ice to inform uncertainty-aware decisions about ice core drilling and other climate research efforts, which are critical to understand the past, and predict future climate.
I have a Bachelor of Science in Management and Technology from the Technical University of Munich, Germany, with a major in chemical engineering. I completed the Graduate Certificate in Data Science (with High Distinction) from the University of Sydney, followed by a Master of Data Science (with High Distinction), also from Sydney.
Ratneel is a DARE PhD candidate studying geophysical modeling at the University of Sydney with a research interest in creating a synergy of optimization and inference methods. Prior to that, he completed his MSc in 2018 at the University of the South Pacific, majoring in Computing Science. In his postgraduate degree, he worked on gradient-based and coevolutionary neural learning to assist in writing his master’s degree research thesis.
In his short career, he has worked with mainly time series forecasting problems with an application to cyclone intensity and path prediction, with a recent interest in working with large-scale climatic modeling methods.
Shuang is a postdoc researcher at the School of Civil and Environmental Engineering at UNSW Sydney. Shuang’s PhD research was understanding HABs in shallow waterbodies. Her PhD was supported by the Nuisance and Harmful Algae Science Practice Partnership with Melbourne Water. She commenced her PhD studies after working at the China Meteorological Bureau on weather forecasting and then working in STAR Water Solutions in Sydney on different projects, including data analysis of Australian stormwater quality.
Shuang has an MEng (Environmental Engineering) from UNSW and BSc (Environmental Science) from China Agricultural University.
Dilani is a graduate of Bachelor of Science in Statistics and Operations Research (SOR) Honours Degree who obtained a G.P.A of 3.83 /4.00 thereby being the second academic performer of SOR batch at University of Peradeniya, Sri Lanka. Previously worked as an Assistant Lecturer in the Data science program at Sri Lanka Institute of Information Technology (SLIIT) and also worked as a temporary Lecturer in University of Kelaniya, Sri Lanka.
Dilani’s research interests are the development of novel Bayesian variable selection for spatiotemporal models in quantile regression, Bayesian Inference, Applied Statistics, and Computational Statistics.
In 2019, Vihanga graduated from the University of Sri Jayewardenepura (USJ), Sri Lanka with First Class Honours in Statistics, securing the third place in her cohort. She also holds an Honours Bachelor’s degree in Quantity Surveying (2018) from Birmingham City University, United Kingdom with a Second Class Upper.
Vihanga’s research interests mainly focus on analysis and forecasting of Time Series and Spatio-Temporal data, Bayesian Inference, Applied Statistics and Computational Statistics. Her honours research project titled ‘Forecasting Bitcoin Price Behaviour subject to Dynamic Volatility Condition’ contributed to an abstract publication at the 40th International Symposium on Forecasting (ISF) in 2020.
Climate change is expected to adversely impact human and natural water systems. In this presentation I will summarise some of our recent research that aims to improve our understanding of these changes using a mix of data-driven and process-based approaches. In our work we leverage the opportunities provided by globally available remotely sensed data as well as in situ data sources to applications such as future streamflow, waterbody health and harmful algal blooms.
Speaker: DARE Chief Investigator, Fiona Johnson
Associate Professor Fiona Johnson is the Director of the Water Research Centre at UNSW and is an academic in the School of Civil and Environmental Engineering. She has over 20 years’ experience in hydrology working as a consultant, for government and in academia.
Fiona’s areas of research and teaching focus on statistical hydrology, particularly with respect to flooding and extreme events and the use of global climate models for climate change assessments of water resources systems. She has a particular interest in solutions to climate and hydrological challenges faced by communities in the Global South.
Many tasks in sciences or engineering require the underlying causal information. Since it is typically expensive and time-consuming to conduct randomised experiments, there has been significant attention towards revealing causal relations through the analysis of purely observational data, commonly known as causal discovery. Over the past few years, with the rapid development of big data, causal discovery is facing great opportunities and challenges.
In this talk, I will first introduce some classical causal discovery methods, including PC algorithm and LiNGAM, which has been successfully applied to the cases without latent variable. However, in complex systems, we typically fail to collect and measure all task-relevant variables. In the second part of the talk, I will focus on causal structure recovery in the presence of latent variables. In particular, I will briefly review some research in this line and introduce our recent work, the latter requires less restrictive assumption and hence can handle more general cases.
Speaker: DARE Chief Investigator, Tongliang Liu
Tongliang Liu is an Associate Professor with the School of Computer Science and The Director of Sydney AI Centre at the University of Sydney. He is broadly interested in the fields of trustworthy machine learning and its interdisciplinary applications, with a particular emphasis on learning with noisy labels, adversarial learning, causal representation learning, transfer learning, unsupervised learning, and statistical deep learning theory.
He has authored and co-authored more than 200 research articles including ICML, NeurIPS, ICLR, CVPR, ICCV, ECCV, AAAI, IJCAI, JMLR, and TPAMI. He is/was a (senior-) meta reviewer for many conferences, such as ICML, NeurIPS, ICLR, UAI, AAAI, IJCAI, and KDD, and was a notable AC for NeurIPS and ICLR. He is a co-Editor-in-Chief for Neural Networks, an Associate Editor of TMLR and ACM Computing Surveys, and is on the Editorial Boards of JMLR and MLJ.
Willem has a PhD in field hydrology from the University of Georgia in the US and Agricultural Engineering undergraduate degree from Wageningen University. He is the leading hydrologist at The University of Sydney and an expert in quantitative Hydrology and Catchment Management and simulation modelling. His main research focus is on sustainable water management to balance climate and human impacts. He has a specific interest in agricultural management and impacts.
Willem combines remote sensing, field data and simulation modelling to develop quantitative tools scaling from the field to the continent. Current projects include Bayesian model optimisation, multi-objective model optimisation, understanding model uncertainty in relation to observed water quality and quantity, and the use of satellite data to improve model structures and predictions. Recent work developed soil moisture predictions for Australia and water accounting tools from the field to the landscape.
Willem has extensive experience working with industry in projects with the Cotton industry, Grains industry, Icon water, and more recently with a European/Australian consortium of SME in the water value chain.
Willem has worked in Indonesia, Uruguay, Mexico and India, delivering capacity building programs and management advice. He is currently advising the Uruguayan Agricultural research institute on future research directions in water.
Please join us for lunch in the foyer following the Symposium.
Can’t join in person? Register here to watch the livestream via Zoom.
THRIVING IN UNCERTAINTY: CHANGE QUANTIFICATION IN NATURAL RESOURCES
You’re invited to the DARE Symposium showcasing work from our multidisciplinary, world-class data scientists!
Join us in welcoming keynote speaker, Professor Scott Sisson, a globally recognised expert in computational and Bayesian statistics. As the driving force behind the UNSW Data Science Hub (uDASH), Professor Sisson leads pioneering research that addresses societal challenges, from environmental issues to industry advancements.
Explore the impactful initiatives of the ARC Training Centre in Data Analytics for Resources and Environments (DARE). Our researchers are leveraging sophisticated data science methodologies to confront Australia’s critical biodiversity, mineral sector, and water security challenges.
See above for the full program – featuring thought-provoking discussions and unique perspectives from the forefront of data science!