MSc in Data Analytics


Computer Science


GDP is the single most important barometer for the health of an economy. It’s an important input into the decision making processes of government, industry and state institutions such as central banks. To be useful as an indicator, GDP estimates need to be both timely and accurate. To meet the needs of users, many national statistical institutes publish early or flash estimates of GDP which are produced within 30 days after the end of a quarter. Given the long lags involved in the data collection processes which feed into GDP estimates, these flash estimates are often largely model based. Within the EU, the models utilised are typically the workhorse models of statistics and time series econometrics such as regression and ARIMA models. This study seeks to assess whether deep learning approaches can be used to improve the accuracy of early estimates in a flash GDP context. To assess this a number of number of LSTM models were trained with extensive hyperparameter tuning with their accuracy evaluated based on common metrics along with walk forward validation on a test set. These results were compared to a similar approach with time series econometric models such as ARIMA, ARIMA with additional explanatory variables and VAR. The study concludes that ARIMA models with explanatory variables provide the most accurate estimates. The study also provides recommendations for the improvement of Ireland’s flash GDP estimates process. The study recommends the use of additional explanatory variables in the context of ARIMA modelling. This recommendation was based on findings from this study and insights into Ireland’s flash estimate processes gained from in-depth interviews with experts.

Date of Award


Document Type

Capstone Project

Resource Type