The WA Branch of the Statistical Society of Australia is pleased to announce that our next meeting focuses on Early Career and Student Statisticians. The two invited speakers are the Lachlan Robinson, recipient of the 2020 WA Branch Honours Scholarship; and Shih Ching Fu, winner of best presentation at the 2020 Young Statisticians Workshop.

Visitors are welcome to attend this branch meeting.

Date: Tuesday, 11 May 2021

Time: 6:00ᴘᴍ
Location: Cheryl Praeger Lecture Theatre, The University of Western Australia

Presentations

Modelling Background Water Density Within a Bayesian Hierarchical Framework

Lachlan Robinson, The University of Western Australia

Background water density governs numerous ocean dynamic processes which in turn impact the global climate and influence decision making within the offshore maritime industry. Consequently, modelling water density variations within a water column and being able to capture prediction uncertainty are of significant interest to the scientific community and industry. In particular, Australia's North-West Shelf is home to large reserves of natural gas and oil; this region contributes significantly to Australia's economy. We present and compare three different models for water density using a dataset from this location. The models include: a double hyperbolic tangent function with a clear physical interpretation, a polynomial and a non-parametric Gaussian process. This is realised within a Bayesian hierarchical framework, allowing for the quantification of uncertainty and probabilistic inference.

My name is Lachlan Robinson and am currently completing my Honours at UWA in Mathematics & Statistics. From an early age I've been very interested in maths and whilst completing my undergraduate in maths and data science, I realised that statistics was my true passion. Being able to analyse data and then use it to draw insightful conclusions to aid others, is not only fascinating to me, but also thoroughly enjoyable. Outside of maths, I am an avid board game enthusiast, enjoy social sports and playing piano in my spare time.

Using Bayesian Networks to identify significant crime events in WA

Shih Ching Fu, Curtin University

The Daily Crime Review, delivered to the Deputy Commissioner of WA Police by a staff officer each morning, comprises a verbal briefing of the significant crime incidents from the past 24 hours. Incidents mentioned at this briefing are judged as significant based on details such as the suspects and victims involved, any presence of firearms, and the perceived threat to the community. Notably, the preparation of this Review is quite onerous since officers must screen hundreds of criminal incidents each morning. There is therefore an opportunity streamline this screening process by applying statistical techniques to crime incident data.

We shall describe a simple Bayesian belief network (BBN) model for quantifying crime incident significance as perceived by the Deputy Commissioner. The model attempts to encapsulate the intuition of the officers who prepare this briefing, i.e., how they weight various factors in their decision to include an incident in that day's Review. The nodes, edges, and conditional probability relations of the BBN are estimated from interviews and workshops with experienced police officers who have routinely prepared these kinds of briefings.

Using data on 22 variables from the WA Police Incident Management System database, our model classified incidents dated between 2 and 22 October 2019 (13,228 reports) achieving an AUC score of 83.4%. Closer examination of false negatives suggests that the addition of variables on family violence, drug seizures, and linked incidents may improve the model's performance.

Shih Ching Fu is a data analyst and applied statistician based part-time at Curtin University. With backgrounds in both Computer Science and Statistics, Shih Ching enjoys the computational side of problem solving as well as introducing people to the loveliness of R.

Refreshments and Dinner

Members and guests are invited to mingle over wine, beer, cider, soft drinks and pizza from 5:30ᴘᴍ. Following the meeting you are invited to dine with fellow attendees at a nearby restaurant. Early career and student statisticians will receive $20 off their meal should they wish to come.

Visitors are welcome to attend this branch meeting.

Meeting directions

The Cheryl Praeger Lecture Theatre is located on the ground floor of the Mathematics building at The University of Western Australia. The theatre entrance is on the outside of the building. See: OpenStreetMaps, Google Maps (points to the adjacent Blakers Lecture Theatre).

Parking is free on the UWA campus after 5:00ᴘᴍ. We suggest using Car Park 18 accessed from Fairway Entry 1.

REMOTE viewing options

For those that cannot attend in-person, the presentations will be streamed live over Zoom. Please register on this page to get the connection details.

For further information please contact the WA Branch Secretary (ssa.wa.secretary@gmail.com).