Irregular Stock Data Prediction Performance Optimisation Based on the Simple Linear Interpolation
Zhenyu Xu
2024
Abstract
This study examines how Simple Linear Interpolation (SLI) affects stock data’s irregular data processing and prediction performance of machine learning models. Using Tesla stock data over ten years, this study cleansed, normalised, and applied SLI methods to reduce missing values and inconsistencies in the data. Then, the performance of the models before and after interpolation was evaluated by constructing various machine learning models, including XGBoost, Random Forest, K Nearest Neighbour (KNN) and Stacked Model. The experimental results suggest that SLI enhance the models' performance, especially the most significant improvement for the stacked model. This suggests that SLI, as a data preprocessing technique, can significantly enhance the model's predictive ability by improving the data's completeness and consistency. However, there are differences in the response of different models to SLI, and the performance enhancement of simple models such as KNN is more limited, suggesting that SLI needs to be carefully selected based on the complexity of the model and the data characteristics when applying SLI. This study provides empirical support for data preprocessing in financial data modelling and highlights the crucial role of data preprocessing in enhancing the performance of machine learning models.
DownloadPaper Citation
in Harvard Style
Xu Z. (2024). Irregular Stock Data Prediction Performance Optimisation Based on the Simple Linear Interpolation. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 398-406. DOI: 10.5220/0013264100004568
in Bibtex Style
@conference{ecai24,
author={Zhenyu Xu},
title={Irregular Stock Data Prediction Performance Optimisation Based on the Simple Linear Interpolation},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={398-406},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013264100004568},
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 - Irregular Stock Data Prediction Performance Optimisation Based on the Simple Linear Interpolation
SN - 978-989-758-726-9
AU - Xu Z.
PY - 2024
SP - 398
EP - 406
DO - 10.5220/0013264100004568
PB - SciTePress