Project Description

GovHack 2018
·       Intro and overview
o   A high performing cross functional taskforce including members from Accenture, sass, and Cloudera has been assembled to aid the government in better understanding the problem of insolvency in Australia. Insolvency costs the Australian economy in excess of $5b per year. 
More specifically this hurts the individuals filing for insolvency, their employees, and the trading partners with outstanding accounts. 
o   The scenario
o   The task
·       Why do we care
o   The cost to the economy
o   The cost to owners
o   The fallout to employees, trading partners and the broader community

For the latest updates on the Microsoft Outlook and O365 issues, go to     Intro and overview
o   A high performing cross functional taskforce including members from Accenture, sass, and Cloudera has been assembled to aid the government in better understanding the problem of insolvency in Australia. Insolvency costs the Australian economy in excess of $5b per year. 
More specifically this hurts the individuals filing for insolvency, their employees, and the trading partners with outstanding accounts. 
o   The scenario
o   The task
·       Why do we care
o   The cost to the economy
o   The cost to owners
o   The fallout to employees, trading partners and the broader community

[‎9/‎09/‎2018 5:00 PM] Briskey, Benjamin:

No Title
·       The approach
o   Data from Australian Financial Security Authority vis data.gov.au
o   Scraped population data from ABS
o   Combined data on SA3 levels
o   Developed machine learning algorithm to predict insolvency risk factors
·       The results
o   Visualisation developed to understand trends and patterns in data
o   QLD has highest insolvency rates by population
o   Top 3 areas are Surfers Paradise, Robina, Noosa
o   Correlation of 24.7% between number of insolvent businesses and unemployment rate – peaking in 2016 at 29.5%
o   Correlation of 14.7% between number of insolvent individuals and unemployment rate – peaking in 2016 at 21.8%
o   Weak negative correlation (12.1% and 6% for businesses and individuals respectively) between median household income and insolvency. 
o   Baselined machine learning model, can be developed further with additional feature engineering or hyper parameter tuning from additional datasets
o   Top insolvent business region per state: Molonglo (ACT), Mount Druitt (NSW) , Casey South (VIC), Springfield- Redbank (QLD), Rockingham (WA), Playford (SA), Brighton (TAS), Palmerstone (NT)
·       Wrap up:
o   Learning Experience
o   Cross functional teams delivering outcomes in short term. 
·       Giving it a go foresees an Australia in which the government can accurately identify at risk individuals and businesses before they reach the financial point of return. Enabling this proactive approach will allow for early intervention, and assistance to be provided where it is needed most.