Data Science Down Under, 8-12 December 2019


Newcastle Beach

The aim of the workshop is to bring together Australian researchers and practitioners as well as key international academics in areas related to data science, e.g., mathematics, statistics and computer science, to discuss recent advancements, share ideas and foster new local and international collaborations. The workshop consists of two integral parts:

  1. Boot camp: Sunday 8th December to Monday 9th December,
  2. Recent advances: Monday 9th December to Thursday 12th December.

We intend to organise ongoing yearly workshops with specific underlying theme for each year's boot camp. In our inaugural year in 2019, the theme of the boot camp is 'Randomised Numerical Linear Algebra' (RandNLA), which is a subject of ongoing intense study. For this, invited speakers who are the pioneers and leading scholars in the field will bring their UoN NewSpace expertise in the area of RandNLA and expose early career researchers and graduate students to related background topics. The aim of the second half of the workshop is to explore not only the interactions between the theme of the boot camp, i.e., RandNLA for 2019, and machine learning, but also a diverse range of topics in data science. For this, speakers from various areas related to the mathematics of data science will present and discuss their cutting-edge research results.

To become acquainted with the topic RandNLA, please see the introductory monograph "Lectures on Randomized Numerical Linear Algebra". The suggested prerequisite for the Boot Camp is an introductory knowledge on probability and linear algebra.

Important Dates

  • Abstract submission closes: Thursday 31st October [deadline extended]
  • Early bird registration closes: Thursday 31st October
  • Registration closes: Friday 6th November
  • Workshop:
    • Boot camp: Sunday 8th December to Monday 9th December
    • Recent advances: Monday 9th December to Thursday 12th December

Other dates will be added as they are decided.


The venue for the workshop is Fort Scratchley Function Centre, Nobbys Road, Newcastle.


The workshop consists of two integral parts:

  1. Boot camp (days 1-2): the boot camp is intended to give the participants an intensive introduction and review of the key topics of the workshop,
  2. Recent advances (days 2-5): during the workshop, the participants will be exposed to the most recent cutting-edge research results not only on the theme of the boot camp, but also a diverse range of topics in the mathematics of data science.

The workshop will involve a dinner, to be held on Tuesday 10th December.

Workshop programme

Abstracts for invited talksDownload abstracts for invited talks

Abstracts for contributed talksDownload abstracts for contributed talks

Sunday, 8 December 2019
9:00-9:15OpeningBoot Camp
9:15-10:00Michael Mahoney
10:00-10:45Michael Mahoney
11:15-12:00Michael Mahoney
12:00-12:45Kenneth L. Clarkson
14:15-15:00Kenneth L. Clarkson
15:00-15:45Kenneth L. Clarkson
16:15-17:00David P. Woodruff
Monday, 9 December 2019
9:00-9:45David P. WoodruffBoot Camp
9:45-10:30David P. Woodruff
11:00-11:50Michael Mahoney, Kenneth L. Clarkson, and David P. WoodruffBoot Camp Wrap-Up
11:50-12:40Michael MahoneyMinimax Experimental Design, Exact Expressions for Double Descent, and Implicit Regularization in RandNLA Algorithms
14:00-14:50David P. WoodruffTowards a Zero-One Law for Column Subset Selection
14:50-15:40Kenneth L. ClarksonQuantum-inspired-inspired Algorithms for Data Analysis
16:10-17:00Michael E. HouleLocal Intrinsic Dimensionality: A Practical Foundation for Dimensionally-aware Data Analysis
Tuesday, 10 December 2019
9:00-9:50Deanna NeedellSimple Approaches to Complicated Data Analysis
10:20-11:10Matt WandStreamlined Variational Inference for Random Effects Models
11:10-12:00Peter TaylorSome Thoughts About a Distributed Solution of the PageRank Equation
13:00-23:00Excursion and Dinner
Wednesday, 11 December 2019
9:00-9:50Kate Smith-MilesParty Tricks with Numerical Linear Algebra and the Quest for Trust
9:50-10:40Kerri MengersenBayesian Statistical Analysis of Large Images
11:10-11:40Vivak PatelOn the Practice of Solving Randomly Sketched Linear Systems
11:40-12:10Daniel AhfockOn Randomised Sketching Algorithms and the Tracy-Widom Law
12:10-12:40Lindon RobertsImproving the Scalability of Model-based Derivative-free Optimization
14:00-14:30Peter EadesLarge Graph Visualisation Using Spectral Sparsification
14:30-15:00Mikhail KamalovSemi-supervised VAE with PageRank
15:00-15:30Zdravko BotevHow Bayes Can Help Frequentist Model Selection
16:00-16:30Scott LindstromSplitting Methods for Signal Recovery
16:30-17:00Sevvandi KandanaarachchiDimension Reduction for Outlier Detection
Thursday, 12 December 2019
9:00-9:30Fred RoostaReproducing Stein Kernel Approach for Correcting Approximate Sampling Algorithms
9:30-10:00Ali EshraghLSAR: Efficient Leverage Score Sampling Algorithm for the Analysis of Big Time Series Data
10:00-10:30Glen Livingston JrARMA Models and Big Data
11:00-11:30Samudra HerathName-like Numbers for Simulating Names in Entity Resolution
11:30-12:00Yang LiuStability Analysis of Newton-MR Under Hessian Perturbations
12:00-12:30Russell TsuchidaRicher Parameter Priors for Infinitely Wide MLPs
14:00-14:30James Juniper‘Unreasonable Effectiveness’ of Machine Learning in Both the Natural Sciences and the Social Sciences?
14:30-15:00Vektor DewantoA Review on Average-reward Reinforcement Learning
15:00-15:30Robert KingSampling Behaviour of L-moment Estimators of the GPD Type Generalised Lambda Distribution


Confirmed invited speakers:

Kenneth ClarksonIBM Research, USA
Michael HouleNational Institute of Informatics, Japan
Michael MahoneyUniversity of California, Berkeley, USA
Kerrie MengersenQueensland University of Technology, Australia
Deanna NeedellUCLA, USA (Teleconference talk)
Joshua RossUniversity of Adelaide, Australia
Kate Smith-MilesUniversity of Melbourne, Australia
Peter TaylorUniversity of Melbourne, Australia
Matt WandUniversity of Technology Sydney, Australia
David WoodruffCarnegie Mellon University, USA


Before 31st October (early bird):

  • Students and female ECRs: $85
  • AMSI/AustMS/ANZIAM members: $170
  • non-AMSI/AustMS/ANZIAM members: $230

After 31st October:

  • Students and female ECRs: $110
  • AMSI/AustMS/ANZIAM members: $230
  • non-AMSI/AustMS/ANZIAM members: $300
Registration fees include attending all sessions, morning/afternoon teas, Hunter Valley tour and the conference dinner. Prices are GST inclusive.

Click here to register for the workshop on our Eventbrite page

AMSI Travel Funding

Higher-degree research students in mathematical sciences and Early Career Researchers (ECRs) at AMSI Member Institutions attending the workshop may apply to receive full or partial travel support through the AMSI Travel Fund. Applicants must, at the time of the event, satisfy one of the following:

For more information on the process and to apply, please visit the AMSI Travel Funding form.

AustMS WIMSIG Awards

Additional travel support may be available to workshop attendees through one of these schemes:

  • AustMS WIMSIG Anne Penfold Street Awards: provides financial support to Australian mathematicians for their caring responsibilities, while they travel for conferences or research visits to collaborators, with approximately four Awards awarded annually. The potential uses of these Awards include, but are not limited to, short-term childcare or professional carers for elderly relatives; and
  • AustMS WIMSIG Cheryl E. Praeger Travel Awards: provides support for Australian female mathematicians to attend conferences or to visit collaborators.

From Newcastle Airport

For those arriving at the Newcastle Airport we recommend taking a taxi to Newcastle. The taxi rank is adjacent to the arrivals area of the terminal. Newcastle Taxis can be contacted directly, free-of-charge, on the dedicated taxi phone located in the arrivals end of the terminal. Alternatively, you can catch the 130 or 131 bus from the Newcastle Airport to the Newcastle Station. From Newcastle Station it is an easy walk to the recommended hotels. For more information and/or to plan your exact trip times see the Sydney Trains or the Port Stephens Coaches timetables.

From Sydney Airport

For those arriving at Sydney Airport we recommend taking the train. For information and trip planning, please visit The trip typically takes 3 hours, though the route is scenic. You should take the T2 Airport Line from Domestic or International Airport Stations to Central Station and then the Central Coast & Newcastle Line.

Local Transportation in Newcastle

Newcastle Taxis: bookings can be made online or by calling 133 300 within Australia. For public transport options, please visit

Accommodation Suggestions

We have recently been advised that a company contacted a couple of invited speakers in regards to booking their hotel accommodation on behalf of the Data Science Down Under workshop. We would like to confirm that the company has no relation to our event or our university. Please do not book with them as we do not know if they are legitimate. If you have any inquiries about your booking, please contact Ms Juliane Turner directly.

There are two main areas with hotels and backpacker accommodation — Newcastle East and Honeysuckle, within easy walking distance of each other and the conference venue. Honeysuckle is a popular area with lots of restaurants with great views across the harbour. Additionally, the website sometimes has good deals.

Newcastle East: Honeysuckle:


Ali Eshragh (Chair)University of Newcastle, Australia
Fred Roosta (Co-chair)University of Queensland, Australia
Ricardo CampelloUniversity of Newcastle, Australia
Elizabeth StojanovskiUniversity of Newcastle, Australia
Natalie ThamwattanaUniversity of Newcastle, Australia


Scientific inquiries: Ali Eshragh ( or Fred Roosta (
Logistical inquiries: Juliane Turner (
Technical support: David Allingham (


Download the workshop poster Download the workshop poster.

Abstract Submission

If you want to present a talk or a poster, please submit your abstract through the following link:

Abstract System

Note that abstract submission is a separate process from registering to attend the conference.

The deadline for abstract submission is 31st October 2019 (extended from the end of August).