Stock Price Prediction Based on LSTM-GBM: Evidence from Haier Smart Home

Zimo Li

2024

Abstract

In the past few years, stock prices forecasting has been a hot topic due to the non-linear relationships and the uncertainty of the stock prices. In the meantime, machine learning, especially deep learning method, has made great progress. More and more studies show that machine learning models might capture features that are hard for traditional methods to observe, which means these models might be more apposite for forecasting time series. A new model named LSTM-GBM is designed to forecast the share prices of Haier Smart Home. The long short-term memory (LSTM) model is responsible for giving a specific prediction while the Geometric Brownian Motion (GBM) model is in charge of adding uncertainties to the predictions. The final prediction path will be generated through a filtrating mechanism, which makes a secondary screening of the two models. In addition, a possible upgrading model named LSTM-GBM-LSTM is proposed which is adding a LSTM model after the filtrating mechanism. This thesis compares the performances of LSTM-GBM model with LSTM model, GBM model and LSTM-GBM-LSTM model. The results indicate that LSTM-GBM has made the best prediction. These results suggest that it is feasible to project the stock price through LSTM-GBM model. Besides, more effort is needed to improve the performance of LSTM-GBM-LSTM model.

Download


Paper Citation


in Harvard Style

Li Z. (2024). Stock Price Prediction Based on LSTM-GBM: Evidence from Haier Smart Home. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 214-221. DOI: 10.5220/0013213200004568


in Bibtex Style

@conference{ecai24,
author={Zimo Li},
title={Stock Price Prediction Based on LSTM-GBM: Evidence from Haier Smart Home},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={214-221},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013213200004568},
isbn={978-989-758-726-9},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI
TI - Stock Price Prediction Based on LSTM-GBM: Evidence from Haier Smart Home
SN - 978-989-758-726-9
AU - Li Z.
PY - 2024
SP - 214
EP - 221
DO - 10.5220/0013213200004568
PB - SciTePress