A long short-term memory algorithm-based approach for univariate time series forecasting with application to GDP forecasting

This work presents a time series forecasting method based on Long Short-Term Memory (LSTM) network, which can be utilized for macroeconomic variable forecasting, like Gross Domestic Product. LSTM is popular in Artificial Neural Networks and an active research topic, however applications are limited. The current focuses one-step ahead forecast, uses Python Keras libraries the implementation. applied to forecast Greek Product accuracy results high comparable ARIMA approach. model we present offers competent approach GDP with traditional statistical approaches. It demonstrates feasibility of further developed parameter tuning application diverse large data sets.