Deakin Applied Machine Learning Lab


Deakin Applied Machine Learning Lab foucsses on doing research on the application of Machine Learning. For example, we work on research projects dealing with the application of Machine Learning in Archeaology, Satellite Communication, Ecology, Call Centers, and many other areas.

We are also doing core research in Machine Learning and Data Science with development of a) cutting-edge models for large quantities of data, b) models for explainabe Artificial Intelligence and c) models with uncertainty quantification, etc.

Browse through some of our projects and get in touch if interested.

 



Research Streams


Satellite Communication AI

This research stream investigates the use of Artificial Intelligence in managing Enterprise satellite Communication network. We are exploring various areas such as anomaly detection, asset management as well as adversarial attacks detection. This project will also investigate the optimization of satcomm network, and how AI can be used to achieve some forms of global optimization. This work is done in partnership with Prof. Jinho Choi and Dr. Jihong Park.

We are seeking applictions from a potential Post-Doc to work with us on a six months funded Level A (Step 6) position.

Network Security AI

This stream of research investigates Artificial Intelligence and Machine Learning methods in cybersecurity. The project is done in consultation with Dr. Mengmeng Ge of RMIT University, Victoria. We are building efficient techniques for detecting botnet attacks, as well as working on optimal deceptive defences (e.g., optimal deployment of decoys in complex networks).

Call Center AI

We explore the application of Artificial Intelligence in call-center environment. We are working on automatically scoring the quality of dialogues between caller and the agent in a call center. The project undertakes formalization of the problem such that an NLP framework can be developed.

 

Rock Art AI

We explore the application of Machine Learning and Artificial Intelligence methods in Archaeology. We are currently working on automatically recognizing images that contain painted rock art. The project is done in consultation with Dr. Andrea Jalandoni of Griffith University, Queensland. We are working on building machine learning models to automatically detect multiple locations of rock art in an image. There are several other applications that need to be investigated such as automatic motif extraction, building a rock-art knowledge graph, etc.

We are seeking applications from prospective Ph.D. candidates as well as visitors in our lab to work on this project.

About the Figure: “Magnificent Gallery, Far North Queensland. Permission to use images from Traditional Owner: Johnny Murison", Photograph: Andrea Jalandoni.

 

Physics Informed Machine Learning

The project is aimed to combine physical equations with deep learning data driven methods to develop surrogate models for structural response under various loading condition. To do so, we intend to build from the concept of physics-informed neural networks (PINN). Here, the Artificial Neural Network (ANN) are trained with known governing equations of the underlying physics, initial and boundary conditions. The architecture will then be used to estimate solutions from limited gathered data.

We are seeking applications from prospective Ph.D. candidates as one fully funded Ph.D. Scholarship is available.

 

Join DAMLL


Current Positions

    Check available positions Here.

Apply for Visiting Scholar

    Fill the EOI to join DAMLL as a visiting scholar EOI Form »

Note for SIT723/724 Students

    If you are interested in taking a research project in DAMLL under SIT723/724 program, you should first fill in the EOI.