SSA Canberra invites you to its October branch meeting, which will feature the shared winners of the 2021 Dennis Trewin prize!

Times (+/- some standard errors): October 26,2021 5:45pm – 6:45pm AEDT 

RSVP: No RSVP will be required for this meeting; please see below for the Zoom link. Any issues, please email ssacanberra@gmail.com.

----------------------------------

Speaker 1: Ali Shojaeipour, University of New England

Topic: Investigate the advantages and existing barriers in biometric identification of Cattle.

Description: Today, due to the increasing demand for livestock products, farmers are encouraged to increase the population of their livestock. Obviously, this will have a direct impact on the management and welfare of animals. The most basic step in managing livestock is how to identify them. Animal identification is always a challenging task, although, with the advancement of technology, novel methods are expected to replace traditional methods of animal identification. This presentation focuses on the key limitations of biometric identification of cattle and ways to deal with them. With a brief look at the studies and research that have been conducted in this field, the required system for use in the practical setting will be discussed. The implemented method and the obtained results are then reviewed to determine whether the use of a computer-based cattle identification method can have a positive impact on livestock management while their welfare is also considered.

Biography: Ali Shojaeipour has a wide range of experiences in the fields of Artificial Intelligence, Animal Biometric Identification, Medical Image Processing, and Robotics. His PhD thesis title is "Biometric Identification of Cattle via Deep Learning in a Few-Shot Learning Context". He is currently working in the school of science and technology at the University of New England (UNE), Australia.

---------------------------- 

Speaker 2: Xian Li, ANU

Topic: Subbagging Variable Selection

Description: With more and more massive datasets being unprecedentedly available, two recent trends are worthy of attention in this field. Firstly, computational methods are shifting towards parallel architectures. Secondly, “subsampled” data analysis framework is becoming prevalent. Subbagging (subsample aggregating) happens to bear these desirable features in big data settings. In this work, we propose a subbagging variable selection method. By combining several subsample estimators, we can obtain the subbagging loss function, which approximates the full sample loss function using a quadratic form. The shrinkage estimation and variable selection can be further conducted based on this subbagging loss function. We theoretically establish estimation consistency and selection consistency for this approach. It is also proved that the resulting estimator possesses the oracle property. However, variance inflation is found in its asymptotic variance compared to the full sample estimator.  A modified BIC-type criterion is further developed specifically to tune the hyperparameter in this method. An extensive numerical study is presented to illustrate the finite sample performance and computational efficiency.

Biography: Xian Li is a 2nd year PhD student in statistics from Research School of Finance, Actuarial Studies and Statistics at the ANU. He has worked on several projects including subbagging estimation and variable selection. His recent research topic is about the development of distributed computing methods for big data.

--------------------------- 

Zoom link:

Join Zoom Meeting
https://anu.zoom.us/j/84613678973?pwd=VWpRWnhFMFJIRlY1ZGVId3BGT3ZaZz09

Meeting ID: 846 1367 8973
Password: 150402
One tap mobile
+61731853730,,84613678973#,,,,0#,,150402# Australia
+61861193900,,84613678973#,,,,0#,,150402# Australia

Dial by your location
        +61 7 3185 3730 Australia
        +61 8 6119 3900 Australia
        +61 8 7150 1149 Australia
        +61 2 8015 6011 Australia
        +61 3 7018 2005 Australia
Meeting ID: 846 1367 8973
Password: 150402
Find your local number: https://anu.zoom.us/u/kekNBGJCeV

Or an H.323/SIP room system:
    Dial: +61262227588 (AUCX)
    or 84613678973@zoom.aarnet.edu.au
    Meeting ID: 84613678973
    H323/SIP Password: 150402


Join by Skype for Business
https://anu.zoom.us/skype/84613678973

Website links: 

https://statsoc.org.au/Canberra-Branch-meetings