Reservoir Computing for Macroeconomic Forecasting with Mixed Frequency Data

Repository for the UKRI funded project Reservoir Computing for Macroeconomic Modelling, Ref: ES/V006347/1.

Project structure:

• python: model code; MIDAS and multi-frequency ESN model libraries; DFM model scripts; Jupyter notebooks with estimation, forecasting and simulations.

• R: data clearning; model forecast restructuring and statistical tests (MCS etc).

• figures: all exported figures used in the paper.

Wherever possible, raw data is also included in the data folder. For details regarding data copyrights and sources, please refer to data/info.txt.

Reference

Published version (Open Access): DOI

Working paper versions: ArXiv, ResearchGate

Citation:

@article{BALLARIN2023,
  title = {Reservoir computing for macroeconomic forecasting with mixed-frequency data},
  journal = {International Journal of Forecasting},
  year = {2023},
  issn = {0169-2070},
  doi = {https://doi.org/10.1016/j.ijforecast.2023.10.009},
  url = {https://www.sciencedirect.com/science/article/pii/S0169207023001085},
  author = {Giovanni Ballarin and Petros Dellaportas and Lyudmila Grigoryeva and Marcel Hirt and Sophie {van Huellen} and Juan-Pablo Ortega},
}