International School in Artificial Intelligence and its Applications in Computer Science (ISAAC)
International School in Artificial Intelligence and its Applications in Computer Science (ISAAC)
Enhancing access to cutting-edge topics in AI
Rated ‘five out of five’ for AI in Australia (ERA National Report), Monash University is proud to host ISAAC, the most exciting graduate school on AI in Australia.
This school gives PhD, Masters and Honours students as well as ECRs across the globe a greater opportunity to engage with state-of-the-art advancements in AI, and computer science more broadly.
During this multi-day academic forum, attendees will enjoy:
- lectures delivered by globally-renowned experts in AI and computer science;
- learning about core theories, tools and techniques in different subfields of AI, such as, machine learning, optimisation, planning, knowledge representation, data science, and more;
- getting insights into innovative research developments and their real-world applications;
- having the opportunity to meet world experts in different areas of AI, and access to a growing network of academics, researchers, and graduate students interested in many aspects of AI.
For ECRs and current PhD students, ISAAC will enrich their education by deep diving into topics ranging from traditional research in AI all the way to new applications in other computer science disciplines. Future PhD students will also benefit from early exposure to graduate research as well as the modern topics and developments in AI and Computer Science that will be explored during this event.
Registration is open!
Register now (registration closes on the 13th of November 2021)
Because ISAAC is fully funded and planned to be an in-person event, it has a limited number of available places.
We encourage all interested participants to apply for registration as soon as possible using the registration link above. Notification about the outcome of applications will be sent soon after the registration deadline.
If you are a Masters or Honours student in Australia who is interested in attending ISAAC, doing a PhD in Computer Science or Artificial Intelligence, and does not live in Melbourne, we have a limited number of scholarships available to cover your:
- travel expenses
What is included depends on where you are based in Australia.
Apply now! (Applications will be received until the 13th of November 2021)
CSIRO and DATA61
Probabilistic Modelling and Reasoning: A Machine Learning Approach. In many machine learning applications it is imperative to build robust systems that allow for the combination of prior knowledge with noisy observational data, update our (posterior) beliefs on the variables of interest and, ultimately, enable intelligent decision making. This process of reasoning under uncertainty given prior knowledge and observations can be carried out under the elegant formalism of Bayesian inference. In this series of talks, I will overview the fundamental principles of Bayesian modelling and modern probabilistic inference with a focus on a class of non-parametric models based on Gaussian process priors. I will describe recent advances in scalable and efficient automated probabilistic reasoning and illustrate how these advances have allowed us to carry out uncertainty quantification in highly complex models, including probabilistic deep learning architectures.
Edwin Bonilla is a Principal Research Scientist and leader of the Machine Learning Research Group at CSIRO’s Data61. He obtained a PhD in Informatics at the University of Edinburgh (UK) in 2008. His expertise is in probabilistic modelling and inference algorithms for the analysis of complex data, in areas such as scalable Bayesian inference, Gaussian processes and multi-task learning. Edwin is known for his work on multi-task learning with Gaussian processes and on compiler optimisation with machine learning. He also developed one of the first approaches to scalable and generic inference in models with Gaussian process priors and ‘black-box’ likelihoods and state-of-the-art inference methods for deep Gaussian process models. Dr Bonilla publishes at the top machine learning venues such as NeurIPS, ICML and AISTATS, and provides academic service to the research community as a reviewer and Area Chair for these conferences. He has received test-of-time paper awards in 2017 and 2019 (at the international symposium on Code generation and optimization). Prior to joining Data61 in 2018, Edwin held research positions at UNSW, NICTA, the Australian National University and The University of Edinburgh. Besides the above areas, his most recent work covers problems in Bayesian optimisation, neural differential equations and graph neural networks.
Curtin University and Australian National University
Responsive AI: Physiological and behavioural analytics for health related outcomes. Human beings leak vast amounts of data into the world, which we can capture using sensors and interpret using AI techniques such as neural networks or deep learning. The aim is to mimic what humans do automatically in interactions when we make strong predictions about each others' intentions and internal states. Such approaches are necessary if we are to ever make truly responsive computer systems which takes the human user into account. My students and I have shown excellent results in health related areas from stress to depression. In this lecture series, I will also discuss responsible AI: a privacy-by-design approach to ameliorate the risks inherent in reliably predicting human internal states.
Tom Gedeon holds the Optus Chair in Artificial Intelligence at Curtin University and is an Honorary Professor of Computer Science at the Australian National University, where he was formerly Deputy Dean and Head of Human Centred Computing. His BSc and PhD are from the University of Western Australia, and Grad Dip Management from UNSW. He is twice a former President of the Asia-Pacific Neural Network Assembly, and former President of the Computing Research and Education Association of Australasia. He is currently a member of the Australian Research Council's College of Experts. He is an associate editor of the IEEE Transactions on Fuzzy Systems, and the INNS/Elsevier journal Neural Networks. Professor Gedeon’s research focuses on responsive AI (mainly neural, deep learning, fuzzy and evolutionary) for human centred computing (wearable physiological signals, fNIRS, thermal, EEG, computer vision) to construct truly responsive computer systems (biometrics and affective computing) and humanly useful information resources (hierarchical and time series knowledge), industrial (mining, defence) and social good (medical, educational) applications.
University of Melbourne
Stress-testing algorithms via Instance Space Analysis. Instance Space Analysis (ISA) is a recently developed methodology to support objective testing of algorithms. Rather than reporting algorithm performance on average across a chosen set of test problems, as is standard practice, ISA offers a more nuanced understanding of the unique strengths and weaknesses of algorithms across different regions of the instance space that may otherwise be hidden on average. It also facilitates objective assessment of any bias in the chosen test instances, and provides guidance about the adequacy of benchmark test suites. Where existing benchmarks are demonstrated to be inadequate to cover the entire instance space, ISA provides guidance about how useful new test instances can be generated. This series of lectures provide an overview of the ISA methodology, and the online software tools that are enabling its worldwide adoption in many disciplines. Case studies will be presented to illustrate the methodology and tools, including optimisation and computer vision applications. Finally, the creation of a mathematically generated art work will be presented as an unexpected outcome of instance space analysis.
Kate Smith-Miles is a Melbourne Laureate Professor of Applied Mathematics in the School of Mathematics and Statistics at The University of Melbourne, and Director of the ARC Industrial Transformation Training Centre for Optimisation Technologies, Integrated Methodologies and Applications (OPTIMA). She is also Associate Dean (Enterprise and Innovation) for the Faculty of Science at The University of Melbourne. Prior to joining The University of Melbourne in September 2017, she was Professor of Applied Mathematics at Monash University, Head of the School of Mathematical Sciences (2009-2014), and inaugural Director of the Monash Academy for Cross & Interdisciplinary Mathematical Applications (MAXIMA) from 2013-2017. Previous roles include President of the Australian Mathematical Society (2016-2018), and membership of the Australian Research Council College of Experts (2017-2019). Kate obtained a B.Sc(Hons) in Mathematics and a Ph.D. in Electrical Engineering, both from The University of Melbourne. Commencing her academic career in 1996, she has published 2 books on neural networks and data mining, and over 280 refereed journal and international conference papers in the areas of neural networks, optimisation, data mining, and various applied mathematics topics. She has supervised 29 PhD students to completion, and has been awarded over AUD$20 million in competitive grants, including 13 Australian Research Council grants and industry awards. She was awarded a Georgina Sweet Australian Laureate Fellowship from the Australian Research Council (2014-2020), enabling her Instance Space Analysis methodology to be expanded into an online tool (MATILDA, Melbourne Algorithm Test Instance Library with Data Analytics). Her awards include the Australian Mathematical Society Medal in 2010 for distinguished research; the EO Tuck Medal from ANZIAM in 2017 for outstanding research and distinguished service; the Ren Potts Medal for outstanding research in the theory and practice of operations research from the Australian Society for Operations Research (ASOR) in 2019; and the Monash University Vice-Chancellor’s Award for Excellence in Postgraduate Supervision in 2012.
Modelling Discrete Optimisation Problems. The world is full of decisions about how to allocate scarce resources to get the most use from them, for example in train scheduling, nurse rostering, or mine planning. This class of problems are called "discrete optimisation problems" since we need to make choices from a (large but) finite set of possibilities. In this lecture series we will teach you how to tackle these problems using high level modelling and sophisticated solving technology. We will introduce the MiniZinc modelling language, developed here at Monash, and used around the world to solve industrial discrete optimisation problems. The lectures will expose common "combinatorial substructures" that reoccur frequently in discrete optimisation problems, and show how by identifying and using these combinatorial substructures we can vastly improve the modelling and solving of discrete optimisation problems. The lectures will include hands-on activities teaching the attendees how to use MiniZinc to solve some interesting discrete optimisation problems.
Peter J. Stuckey is a Professor in the Department of Data Science and Artificial Intelligence in the Faculty of Information Technology at Monash University. Peter Stuckey is a Fellow of the AAAI and a pioneer in constraint programming, the science of modelling and solving discrete optimisation problems. His research interests include: discrete optimisation; programming languages, in particular declarative programming languages; constraint solving algorithms; path finding; bioinformatics; and constraint-based graphics. He enjoys problem solving in any area, having publications in, e.g., databases, election science, system security, and timetabling, and working with companies such as Oracle and Rio Tinto on problems that interest them.
Roles for Natural Language Processing in Biomedicine. By some estimates, up to 70% of the electronic health record consists of free text documents. In this lecture series, I will describe why it is important that we make the effort to take advantage of this complex data source for clinical purposes, and how that translates into a number of challenges that can be addressed with natural language processing (NLP). In addition, I will describe how NLP can be leveraged for medical evidence synthesis, and deeper understanding of molecular biology, based on analysis of the scientific literature. I aim to paint a picture of the broader opportunities that exist for data-based decision making in biomedicine.
Professor Karin Verspoor is Executive Dean of the School of Computing Technologies at RMIT University in Melbourne, Australia and a Fellow of the Australasian Institute of Digital Health. Karin’s research primarily focuses on the use of artificial intelligence methods to enable biological discovery and clinical decision support, through extraction information from clinical texts and the biomedical literature and machine learning-based modelling. Karin held previous posts as Director of Health Technologies and Deputy Head of the School of Computing and Information Systems at the University of Melbourne, as the Scientific Director of Health and Life Sciences at NICTA Victoria Research Laboratory, in the Computational Bioscience program at the University of Colorado School of Medicine, and at Los Alamos National Laboratory. She also spent 5 years in start-ups during the US Tech bubble, where she helped design an early artificial intelligence system.
Reasoning with Knowledge Graphs. Knowledge graphs have become a promising modelling tool in a wide variety of applications such as intelligent Web search, question answering, in-context advertising, social media mining, and biomedicine. A knowledge graph is a semantic network for modelling entities (including properties) and their relations in an application domain. This series of tutorials will focus on reasoning and its applications in knowledge graphs. We will provide a brief history of knowledge graphs, introduce major reasoning forms and algorithms, and highlight some practical applications in knowledge acquisition, query answering and question answering using knowledge graphs. Finally, we will discuss some future research directions for knowledge graphs.
Kewen Wang is a Professor in the School of Information and Communication Technology and the Program Leader of AI and Semantic Technologies (AIST) in the Institute for Intelligent and Integrated Systems at Griffith University. His research interests include Knowledge Representation (specifically, answer set programming and description logics), Knowledge Graphs and Ontologies. He has been publishing in the foremost conferences and journals in his area. He is currently an Associate Editor/Area Editor of the Journal of Web Semantics and an Area Chair of AAAI-2022. He is regularly a (senior) member of the Program Committee of major conferences in AI such as AAAI, IJCAI, KR and TheWebConf.
International Speakers giving online plenary talks
University of Oxford
Introduction to Cake Cutting. We will consider the setting whether there is a single divisible item (usually referred to as a "cake") to be divided among multiple agents, who all have their own valuations for different parts of the cake. We will discuss several notions of fairness as well as algorithms for finding fair allocations. No background in fair division is expected.
Edith Elkind is a Professor of Computer Science at the University of Oxford. She works in algorithmic game theory, with a focus on voting theory, coalition formation and fair division. She is a EurAI fellow and has published over 100 papers in leading conferences in AI and algorithmic game theory, and served as a chair of AAMAS, EC and COMSOC as well as an editorial board member of AI Journal, Journal of AI Research, Transactions on Economics and Computation, Social Choice and Welfare and several other journals.
NASA Ames Research Center
SafeDNN: Understanding and Verifying Neural Networks. The SafeDNN project at NASA Ames explores analysis techniques and tools to ensure that systems that use Deep Neural Networks (DNN) are safe, robust and interpretable. Research directions we are pursuing include: symbolic execution for DNN analysis, label-guided clustering to automatically identify input regions that are robust, parallel and compositional approaches to improve formal SMT-based verification, property inference and automated program repair for DNNs, adversarial training and detection, probabilistic reasoning for DNNs. In this talk, I will highlight some of the recent research advances from SafeDNN.
Corina Pasareanu works in software engineering at NASA Ames, in the Robust Software Engineering group. She is employed by KBR, where she is Technical Professional Leader - Data Science. She is also affiliated with Carnegie Mellon University (CMU), CMU CyLab, CMU Silicon Valley, and CMU ECE. Her research interests include: model checking and automated testing, compositional verification, model-based development, probabilistic software analysis, autonomy and security. In her work, she investigates how to use learning and abstraction techniques for automating assume-guarantee style compositional verification, abstraction and symbolic execution with applications to test input generation and error detection, modeling and analysis techniques for multiple statechart formalisms used in the development of NASA systems, probabilistic techniques for the analysis of software components placed in stochastic environments, and planning for safe surface operations for autonomous vehicles, and the analysis of deep neural networks. Corina has a PhD from Kansas State University, numerous scientific awards, and is frequently invited to deliver plenary talks at major academic conferences in Computer Science around the world.
Program and topics
Run across five days, ISAAC will involve a registration period, a welcome event, a number of events organised by the Faculty of Information Technology (FIT events), and a variety of teaching sessions (18 seminars delivered by our 6 invited speakers, with each main speaker delivering 3 seminars), ranging from introductory, intermediate and advanced, that fall within three main tracks: Foundations, Logic, and Automated Reasoning; Machine Learning and Data Science; AI for Systems' Quality and Design. The school will also feature a number of online plenary talks given by experts in AI and Computer Science around the world.
|9.00-10.00||Plenary talk||FIT event||Plenary talk||Seminar 16|
|10.30-11.30||Seminar 1||Seminar 6||Seminar 11||Seminar 17|
|11.30-12.30||Seminar 2||Seminar 7||Seminar 12||Seminar 18|
|14.00-15.00||FIT event||Seminar 3||Seminar 8||Seminar 13|
|15.30-16.30||FIT event||Seminar 4||Seminar 9||Seminar 14|
|16.30-17.30||FIT event||Seminar 5||Seminar 10||Seminar 15|
ISAAC is organised by the Faculty of Information Technology: