This repository contains the code for the paper "Hybrid ARDL-MIDAS-Transformer Time-Series Regressions for Multi-Topic Crypto Market Sentiment Driven by Price and Technology Factors", by Ioannis Chalkiadakis, Prof. Gareth W. Peters and Dr. Matthew Ames.

Please cite the paper as follows:

Chalkiadakis, I., Peters, G.W. & Ames, M. Hybrid ARDL-MIDAS-Transformer time-series regressions for multi-topic crypto market sentiment driven by price and technology factors. Digit Finance 5, 295–365 (2023). https://doi.org/10.1007/s42521-023-00079-9

The model fitting code in R is in folder src/, while the data that were used in the studies are included in the folder data/.

Please make sure to adjust the path names in some of the scripts if necessary.

Tested with R v3.6.1.

Copyright @ Authors August 2021