Forecast and Analysis of the Fifth Trading Day Yield based on the Last 35 Trading days-Based on Long Short-Term Memory Methord
Journal of Innovation and Development · 15.04.2026 · 15 (2) · 80-88
DOI: 10.54097/623d4z05 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Innovation and Development, online erschienen am 15.04.2026. Verfasst von Yiyang Luo. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Throughout the past two decades, the China Securities Index 300 (CSI 300) Index has undergone a frequency of extreme fluctuations. This renders traditional prediction models unable to precisely capture its dynamic changes. The existing GARCH and VAR models can only partially illustrate the statistical traits of the CSI 300 Index. They cannot simultaneously account for the sharp peaks and heavy tails, long - term memory, and the 5 - day operation cycle generated by the T + 1 trading system exclusive to the A - share market. The lightweight Long - Short Term Memory prediction framework contrived in this study uses only a 35×5 OHLCV matrix as input, without bringing in external material. It gets rapid training via a basic double - layer Long Short-Term Memory (LSTM) (32, 16) - Dense (1) structure, aiming to provide a low - cost and useful prediction tool. The double - layer LSTM configured with 32 - 16 neurons gets an ideal balance in parameter quantity, training efficiency, and prediction accuracy.It confirms the principle of complexity - data scale consistency and guarantees the model functions smoothly and profitably.It validates the principle of complexity - data scale consistency and guarantees the model functions.
Zitieren
APA
Luo, Y. (2026). Forecast and Analysis of the Fifth Trading Day Yield based on the Last 35 Trading days-Based on Long Short-Term Memory Methord. Journal of Innovation and Development, 15(2), 80-88. https://doi.org/10.54097/623d4z05
Vancouver
Luo Y. Forecast and Analysis of the Fifth Trading Day Yield based on the Last 35 Trading days-Based on Long Short-Term Memory Methord. Journal of Innovation and Development. 2026;15(2):80-88. doi:10.54097/623d4z05
BibTeX
@article{Luo2026,
author = {Luo, Yiyang},
title = {Forecast and Analysis of the Fifth Trading Day Yield based on the Last 35 Trading days-Based on Long Short-Term Memory Methord},
journal = {Journal of Innovation and Development},
year = {2026},
volume = {15},
number = {2},
pages = {80--88},
doi = {10.54097/623d4z05}
}
EUVIMED