The capstone

All MSc business analytics students carry out a substantial research project called the capstone. As project sponsor, SAS proposed two projects for which we will provide the data and access to the latest SAS analytical software. Both projects will focus on using the Just Eat DublinBikes data accessible via a web API provided by JCDecaux under an open data license agreement. In Project 1, the data will be mined for insights and a dashboard showing a live view of the Dublin Bikes situation will be created. Models predicting supply and demand per station will be created and tested, based on historical data. As a stretch goal, external data sets on special events, weather and other influential variables will be incorporated.

In Project 2, optimisation models will be created for the redistribution of bikes from full to empty stations. Initially, this will work on the simplest possible predictive model. For example, models will predict that demand at a point in time at each station is equal to that at the same station exactly one week earlier. The goal will then be to create a redistribution plan that minimises driving time. It will also maximise the availability of bikes to users, taking into account users’ proximity to multiple stations, probability of empty stations, and other real-world factors. Our four students from UCD who picked these projects were invited into the SAS offices for training and support throughout the project timeline.

 

Left to right: Joseph Kneen (the author), Manh Tuan Nguyen (UCD), Karina Cortinas (SAS), Tasha Brahpunath (UCD), Ben Lyu (UCD) and Robert Pauley (UCD).

The project teams get to grips with using the latest SAS technology, SAS Viya.