Big Data Official Statistic Competition 

As a member of the Warwick Business School (WBS) University of Warwick Forecast Team, I took part in the Big Data Official Statistic Competition organised by Eurostat. Other members as are Doctoral Researchers Danilo Cascaldi-Garcia and Ingomar Krohn as well as Professors Ana Beatriz Galvão and Anthony Garratt from WBS Economic Modelling and Forecasting group.

Among the approaches we used, Unobserved Component Stochastic Volatility (UC-SV) models performed exceptionally well especially when forecasting price indices excluding energy prices for several European countries. In this category, UC-SV models outperformed all other approaches. We also found that mixed models provide an excellent non-judgemental approach.

 

For the final report and a collection of replication codes see:

https://ec.europa.eu/eurostat/cros/content/bdcomp-participant-5-methodological-information_en

 

For the official press release by Warwick Business School see:

http://www.wbs.ac.uk/news/phd-students-win-eurostat-forecasting-competition/