Project Abstract:

The current FT Staff Rosterer is plagued by complicated settings and confusing interface. Deployment of the software often requires special consultancy, and it is hard for the users to make adjustment afterwards. A new artificial intelligence approached is devised to tackle this problem, whereby the system automatically figures out the adjustment of settings from a spectrum of desirable and undesirable schedules provided by the users. The automatic tuning wizard greatly enhances the user experience of FT Staff Rosterer by combining the latest development from operations research and machine learning synergistically.