From our Competition Chair, Tony Nolan:
If you are a current university student then we encourage you to participate in the AusDM 2018 student challenge. If you do win, then you will be given a speaking opportunity at the conference, providing you have registered correctly by the due date. These instructions about registration will be sent to you a few days before the close of registration. My email faddress is email@example.com.
Please register your team name, team membership, and an image of your student ID to firstname.lastname@example.org
The challenge is simple, but the task is complex. To answer the challenge you just need to email me a two page summary of your findings and a copy of your code. Entries are judged on the effectiveness of the methods you used, the accuracy of your results, and on your expression, report style, and the uniqueness of your approach.
The task is to remove 300 post codes from the groups field. The removed postcodes must be from at least 10 groupings, where the frequency of those groups must be 3 or above. No more than 3 can be removed from any group, across all time periods, You mush then design a clustering system to assign those removed post codes back to their original groups and time periods.
An explanation of the groupings is contained within the spreadsheet on
All entries must be received to my email address by midnight October 10th.
Thank you to the Australian Taxation Office from whom the dataset has been provided under Creative Commons Attribution 3.0 Australia.