Predictive Data Analytics
Traditional data management tools are struggling to process the flood of data generated in the digital world.
Extracting business value requires a new approach to make it easier to collect, store, analyse, visualise and dynamically act on data from separate ‘silos’ including:
- internet browsing history and mobile application data
- customer databases and machine logs
- financial transactions
- third party sources (e.g. ABS) and open data.
Data61 researchers make sense of vast amounts of unstructured data to drive business value through:
- utilising Bayesian techniques to predict business outcomes with associated confidence levels
- identifying changing customer behaviours early
- detecting potential business operation failure
- detecting anomalies, potential fraud and quantifying business risks.
Impact
As one of Australia’s largest data science research teams, the researchers have developed a number of Big Data solutions for a growing list of high profile organisations including:
- data workbenches for ingestion and data management
- natural language processing modules
- data de-identification approaches
- privacy preserving techniques
- recommendation engines
- embedded analytics
- spatio-temporal modelling
- visualisation platforms.
Previous post:
Coming up next: