The repo for the LOB Recreation Model (LOBRM) that predicts deep level volume in the LOB from TAQ data, described in the paper PKDD21 AAAI21. Please note the difference in results between chronological and non-chronological training. There might be minor deviations between the paper and the implementation released that do not effectively influence the conclusions.
Please cite:
@inproceedings{shi2021lob,
title={The lob recreation model: Predicting the limit order book from taq history using an ordinary differential equation recurrent neural network},
author={Shi, Zijian and Chen, Yu and Cartlidge, John},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
volume={35},
number={1},
pages={548--556},
year={2021}
}
@inproceedings{shi2021limit,
title={The Limit Order Book Recreation Model (LOBRM): An Extended Analysis},
author={Shi, Zijian and Cartlidge, John},
booktitle={Machine Learning and Knowledge Discovery in Databases. Applied Data Science Track: European Conference, ECML PKDD 2021, Bilbao, Spain, September 13--17, 2021, Proceedings, Part IV 21},
pages={204--220},
year={2021},
organization={Springer}
}