Visualisation and statistical modelling techniques for the management of inventory stock levels

This paper describes the investigations conducted in a Mathematics-in-Industry Study Group project from the Australian meeting at Queensland University of Technology in 2015. This concerned the management of stock levels of raw materials used to construct aortic stents. The approaches used included network visualisation, classification and regression trees, and time series modelling. This work will be of general interest to those who are managing stock levels in a highly volatile context. The methods applied show that there is potential value in taking a statistical approach to understand and make decisions within such volatility. The work provides a basis for developing more advanced statistical approaches for specific inventory problems.