Adaptive Normalization for Forecasting Limit Order Book Data Using Convolutional Neural Networks
Tutkimustuotos › › vertaisarvioitu
Yksityiskohdat
| Alkuperäiskieli | Englanti |
|---|---|
| Otsikko | 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings |
| Kustantaja | IEEE |
| Sivut | 1713-1717 |
| Sivumäärä | 5 |
| ISBN (elektroninen) | 9781509066315 |
| DOI - pysyväislinkit | |
| Tila | Julkaistu - 1 toukokuuta 2020 |
| OKM-julkaisutyyppi | A4 Artikkeli konferenssijulkaisussa |
| Tapahtuma | IEEE International Conference on Acoustics, Speech and Signal Processing - Barcelona, Espanja Kesto: 4 toukokuuta 2020 → 8 toukokuuta 2020 |
Julkaisusarja
| Nimi | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings |
|---|---|
| Vuosikerta | 2020-May |
| ISSN (painettu) | 1520-6149 |
Conference
| Conference | IEEE International Conference on Acoustics, Speech and Signal Processing |
|---|---|
| Maa | Espanja |
| Kaupunki | Barcelona |
| Ajanjakso | 4/05/20 → 8/05/20 |
Tiivistelmä
Deep learning models are capable of achieving state-of-the-art performance on a wide range of time series analysis tasks. However, their performance crucially depends on the employed normalization scheme, while they are usually unable to efficiently handle non-stationary features without first appropriately pre-processing them. These limitations impact the performance of deep learning models, especially when used for forecasting financial time series, due to their non-stationary and multimodal nature. In this paper we propose a data-driven adaptive normalization layer which is capable of learning the most appropriate normalization scheme that should be applied on the data. To this end, the proposed method first identifies the distribution from which the data were generated and then it dynamically shifts and scales them in order to facilitate the task at hand. The proposed nor-malization scheme is fully differentiable and it is trained in an end-to-end fashion along with the rest of the parameters of the model. The proposed method leads to significant performance improvements over several competitive normalization approaches, as demonstrated using a large-scale limit order book dataset.