Analytics in Decentralised Artificial Intelligence Enabled Autonomy

Data mining, analytics, and statistics are normally taught at universities, assuming data can be centralised one way or another. Researchers have been investigating decentralisation of analytics with significant advances made in areas such as data mining, federated learning, learning at the edge, and so forth. However, little attention is given to the 'agent', who is the data custodian? What is the purpose for collecting the data? What can we get out of the data? When? Where? Why? How? In this talk, I will polarise the analytics on the 'agent' and 'uncertainty' in decentralised artificial intelligence. While the presentation will shed light on the earlier and current work of my group at UNSW Canberra, the take-home message will focus more about the future, where we are heading, and opportunities we would love to collaborate on with other Australian teams.

Dr. Hussein Abbass is a full professor with the School of Systems and Computing, University of New South Wales, Canberra. He is a Fellow of the Institute of Electrical and Electronics Engineering (IEEE) USA, a Fellow of the Australian Computer Society, a Fellow of the UK Operational Research Society, a Fellow of the Australian Institute of Managers and Leaders, and a Graduate Member of the Australian Institute of Company Directors. Hussein was the National President (2016-2019) for the Australian Society for Operations Research, the Vice-President for Technical Activities (2016-2019) for the IEEE Computational Intelligence Society, and an ExCom and AdCom member (2016-2019) of the IEEE Computational Intelligence Society. Hussein is a Distinguished Lecturer for the IEEE Computational Intelligence Society and the Founding Editor-in-Chief of the IEEE Transactions on Artificial Intelligence. Hussein is the chair of the IEEE Conference on AI Steering Committee, the incoming chair of the IEEE Frank Rosenblatt Award committee (equivalent to the technical medal in computational intelligence) and is the vice-chair for the Working Group on the IEEE P7018 Standard for Security and Trustworthiness Requirements in Generative Pretrained Artificial Intelligence (AI) Models. Hussein is a UAV pilot and a mental health first-aid officer and has completed various executive professional development training. Following ten years in industry and academia, in 2000, he joined the University of New South Wales campus in Canberra (UNSW-Canberra) at the Australian Defence Force Academy. He has been a full professor since 2007 and has served in various university leadership roles. His current research focuses on trusted quantum-enabled human-AI-swarm teaming systems and distributed and trusted machine learning and machine education systems and algorithms.

Data Science for Large Models

Large Models (LM) has made significant progress and found wide application in various fields, like LLM for question answering. However, the success and efficiency of LM models depend on proper data management. Training a LM is challenging without labeled data, and efficiency is hindered by large datasets, complex models, and numerous hyperparameters. Lack of validation and explanation limits model applicability. In this talk, I will discuss three crucial issues in data science for large models: 1) effective data preparation for LM, including data selection; 2) LM training optimization, involving computation graph optimization; and 3) the importance of model explanation for robustness and transparency. I will conclude by highlighting future research directions on data science for large Models.

Lei Chen is a chair professor in the data science and analytic thrust at HKUST (GZ), Fellow of the IEEE, and a Distinguished Member of the ACM. Currently, Prof. Chen serves as the dean of information hub, the director of Big Data Institute at HKUST (GZ). Prof. Chen’s research interests include Data-driven AI, Big Data Analytics, Metaverse, knowledge graphs, blockchains, data privacy, crowdsourcing, spatial and temporal databases and probabilistic databases. He received his BS degree in computer science and engineering from Tianjin University, Tianjin, China, MA degree from Asian Institute of Technology, Bangkok, Thailand, and PhD in computer science from the University of Waterloo, Canada. Prof. Chen received the SIGMOD Test-of-Time Award in 2015, Best research paper award in VLDB 2022.The system developed by Prof. Chen’s team won the excellent demonstration award in VLDB 2014. Prof. Chen had served as VLDB 2019 PC Co-chair. Currently, Prof. Chen serves as Editor-in-chief of IEEE Transaction on Data and Knowledge Engineering and General Co-Chair of VLDB 2024.

Fortifying Graph Neural Networks in Robustness, Anomaly Detection and Continual Learning

Graphs are ubiquitous and widely used to model complex relationships between instances across various domains, such as social networks, biology, and chemistry. Due to the great capacity of modeling graph data, Graph Neural Networks (GNNs) have been extensively utilized for graph learning. Despite achieving remarkable achievement, GNNs have raised several concerns in real applications, including vulnerability to adversarial attacks, sensitivity to anomalous data, and lack of generality to continually expanded graphs. For example, attackers can fool the GNNs into giving the outcome they desire with unnoticeable perturbation on the training graph and the performance of GNNs would be significantly deteriorated with the existence of anomalous graph data. To address these concerns, we aim to fortify GNNs in robustness, anomaly detection and continual learning. Specifically, by developing novel defense mechanisms, we aim to improve the robustness of GNNs against attacks and noises. Moreover, to prevent the graph data from being contaminated by the anomaly, graph anomaly detection aims to identify abnormal graphs that exhibit deviant structures and node attributes compared to the majority in a graph set. Lastly, to accommodate the continually expanding graphs and enhance the effectiveness of GNNs, graph continual learning would continually adapt GNNs to the expanded graph of the current task while maintaining the performance over the graph of previous tasks.

Dr. Ling Chen is a Professor in the School of Computer Science at UTS, Sydney, Australia. She received her PhD in Computer Engineering from Nanyang Technological University (NTU), Singapore, and undertook postdoctoral training at Leibniz University Hannover (L3S Research Centre), Germany. As the Deputy Head of School (Research) for the School of Computer Science, Ling is leading major research & development activities across different disciplines and research institutes/centres within the school. Ling also leads the Data Science and Knowledge Discovery Laboratory (The DSKD Lab) within the Australian Artificial Intelligence Institute (AAII) at UTS. Ling’s research interests mainly include (i) discovering regularities (e.g., patterns) and irregularities (e.g., outliers or novelties) from various types of data (e.g., structured/unstructured data, single-modal/multimodal data, and static/dynamic data etc.); (ii) data representation learning, including both hash-based and learning-based methods for graph structured data and spatio-temporal data; (iii) social media and social network mining, including event detection, information diffusion modelling and user profiling for recommendation; (iv) dialogue and interactive systems, including reinforcement learning (for POMDP) and continual learning. Ling has secured multiple competitive research grants, including ARC DP/LP/LIEF. Ling’s research has also been recognised and funded by the industry, including Facebook and TPG Telecom. Ling is an Editorial Board member for the IEEE Journal of Social Computing, and the Elsevier Journal of Data and Knowledge Engineering.

Generative Data Mining: Shaping the Future of Adaptive Intelligence in Complex Data Ecosystems

This keynote will explore the frontier of generative data mining as it evolves from a tool for pattern discovery into a foundational technology for creating synthetic data, predicting future states, and enhancing adaptive intelligence. Focusing on applications in areas like predictive finance, healthcare diagnostics, and smart cities, the talk will highlight how generative models can transform static data assets into dynamic, interactive resources. By delving into advancements in generative adversarial networks (GANs), probabilistic models, and reinforcement learning, this keynote will illuminate how these methods enable systems to not only interpret but also generate meaningful scenarios that enhance forecasting and decision-making capabilities in real time. Emphasis will also be placed on the ethical frameworks and technical strategies needed to ensure generative models operate within secure, transparent, and bias-aware ecosystems.

Prof. Amin Beheshti is a Full Professor of Data Science at Macquarie University, and an Adjunct Professor of Computer Science at UNSW Sydney, Australia. Amin is the founder and director of the Centre for Applied Artificial Intelligence, the head of the Data Science Lab, and the founder of the Big Data Society at Macquarie University, Sydney, Australia. Amin completed his PhD and Postdoc in Computer Science and Engineering at UNSW Sydney, and holds a Master's and Bachelor's degree in Computer Science, both with First Class Honours. Before starting his PhD in 2009, Amin had over a decade of industry experience as a founder and CEO, consultant, and Solution Architect in national and international organizations. Alongside his teaching activities, Amin has made significant contributions to research projects and successfully secured 50+ research projects (Over $38 million in Research Funding). Amin received Prestigious Awards, including Excellence Award (Macquarie University, 2023), Excellence in Research Innovation, Partnership Entrepreneurship (Macquarie University, 2022), National Security Impact Award (D2D CRC, 2016 and 2017), Recognition Award (D2D CRC, 2016 and 2017), Australian Postgraduate Award (APA 2009-2012), and several Best Paper awards. In 2021, due to his outstanding performance, Amin was promoted from Senior Lecturer to Full Professor at Macquarie University. As a distinguished researcher in Data and AI Science, Amin has been invited to serve as a Keynote Speaker, General-Chair, PC-Chair, Organisation-Chair, and program committee member of top international conferences. He is also a leading author of several authored books in data, social, and process analytics, co-authored with other high-profile researchers. Amin was named a finalist in the prestigious Australian AI Awards 2024 in three categories: AI Academic / Researcher of the Year, AI Leader of the Year – Enterprise, and AI Rising Star of the Year – Enterprise. Amin has been invited to serve as a Distinguished Jury member for prestigious awards, including the "Aegis Graham Bell Award, recognizing his leadership in AI and significant contributions and involvement in commercialization efforts.

Harnessing Tabular Data Lakes: A Systematic Guide to Discovering and Assembling Data

Data lakes have emerged as vital repositories for storing vast quantities of heterogeneous data, presenting immense opportunities as well as significant challenges for data-driven research and applications. This talk introduces a systematic guide to effectively harnessing tabular data in data lakes, focusing on three key tasks: 1) Dataset Discovery – identifying relevant datasets that align with various user intents and inputs; 2) Dataset-Level Assemblage – assembling the discovered datasets into a unified and comprehensive resource that meets various user requirements; 3) Data Points-Level Assemblage – optimizing the selection of data points from the assembled dataset, curating a subset most effective for typical downstream tasks such as machine learning model training. By addressing these tasks, our guided framework transforms fragmented raw data into high-quality, application-ready datasets. The talk will cover problem formulations, challenges, and methodologies involved, and will highlight open questions where effective and efficient data preparation is crucial. Ultimately, we aim to explore the potential for developing an intelligent, personalized data preparation agent to automate and optimize these processes for real-world applications.

Prof. Zhifeng Bao leads the Big Data and Database Group at RMIT University and is an Honorary Senior Fellow at The University of Melbourne. In the past he co-directed the RMIT Center of Information Discovery and Data Analytics. He obtained his PhD in Computer Science from National University of Singapore and received the Best PhD Thesis Award. His recent research focuses on data management and governance, particularly in DB4AI and AI4DB. In DB4AI, he investigates how to identify suitable datasets, uncover hidden relationships, tackle data quality issues, and meet diverse user needs. In AI4DB, he studies how machine learning can optimize database operations, including index selection, query optimization, and cardinality estimation for both low- and high-dimensional data. He has received several honors, including the Australasian Research Council Future Fellowship, the Computing Research and Education Association of Australasia (CORE) Award for Outstanding Research, the Google Faculty Research Awards, and Best Paper Award Runner-up at KDD’19. He is the PC Co-chair of full paper track at CIKM’24 and has served as the Associate Editor of PVLDB, SIGMOD, and ICDE. He also chairs the Data Management and Data Science field for the CORE 2026 conference ranking committee. In addition to academic work, he provides consultancy to various organizations, including the City of Melbourne on its Smart City Project and the Victoria Department of Health and Human Services on data quality initiatives.

Mathematical Artificial Intelligence

Artificial Intelligence is a leading topic in both academia and industry, and explainable AI (XAI) is a critical and hot topic of the field. However, we noticed that XAI is not the core of the business, and we expect deterministic AI models. In this talk, we will report the current landscape of XAI, and then introduce the next stage after XAI – Mathematical AI (MAI). We will mainly present how the mathematical tools, such as differential geometry and group theory, are used build deterministic AI models. At the end of the talk, we will present some cases of the application of MAI. We hope the talk will shed light on the promising field for interested audience.

Dr. Shui Yu is a Professor of School of Computer Science, University of Technology Sydney, Australia. His research interest includes Cybersecurity, Network Science, Big Data, and Mathematical Modelling. He has published seven monographs and edited two books, more than 600 technical papers at different venues, such as IEEE TDSC, TPDS, TC, TIFS, TMC, TKDE, TETC, ToN, and INFOCOM. His current h-index is 78. Professor Yu promoted the research field of networking for big data since 2013, and his research outputs have been widely adopted by industrial systems, such as Amazon cloud security. He is currently serving the editorial boards of IEEE Communications Surveys and Tutorials (Area Editor) and IEEE Internet of Things Journal (Editor). He is a Distinguished Visitor of IEEE Computer Society, and an elected member of Board of Governors of IEEE VTS and IEEE ComSoc, respectively. He is a member of ACM and AAAS, and a Fellow of IEEE.

The Amazing Power of Open Models

Foundation models and generative AI are now having a major impact across many industries, and those changes are happening throughout the entire business stack, from logistics and worker productivity, to consumer-facing products. Larger, closed models continue to push the limits and perform best-in-class. But open models have some amazing power too, and with great power comes great responsibility. What can we do as an industry to continue to support open models? How can we ensure that AI is being deployed responsibly?

Over 10 years of industry experience, including leading engineering teams at Google and deploying software to billions of users as part of Windows and Android. Experienced in technical outreach and partnerships, responsible AI development, and foundation models as part of the Responsible AI team at Google.