Prof Kerrie Mengersen
Bayesian statistical methods for modelling and data analysis are now endemic in a very wide range of fields. The Bayesian paradigm has a number of appealing features, including the ability to describe complex data structures, characterise uncertainty, and provide comprehensive estimates of parameter values, comparative assessments, probabilities and risk. There is also great benefit and interest in using the Bayesian framework as a mechanism for merging statistical and mathematical models and methods, and smart computational algorithms. This is particularly relevant in the context of big data, which we define as "inconveniently large" data. In this presentation I will draw on some of our experiences in applying Bayesian statistical methods to big data problems in health, environment and industry, with a focus on the nexus between stats, maths and computation. The discussion will be cast in the form of ‘grand challenges’ which involve the development of appropriate theory, models and computational algorithms to address the applied problems.