Journal Browser
Search
Missing Value Filling Research Based on Ensemble Learning
Jianqiao Sun
International Journal of Mathematics and Systems Science 2024, 7(3); https://doi.org/10.18686/ijmss.v7i3.5058
Submitted:06 Mar 2024
Accepted:06 Mar 2024
Published:06 Mar 2024
Abstract
This paper studies missing value filling and compares the filling effects of five methods: Mean, KNN, Random Forest, GBDT, and Stacking under different missing proportions, proving the superiority of ensemble learning algorithms in filling performance when multiple feature values are missing. Then the missing value filling method of KNN+integrated learning is proposed to further improve the filling performance.
References
[1] Zheng Zhiquan, Wang Mengmeng, Tian Weiqi. Research on Missing data filling based on weighted K-nearest neighbor algorithm
[J]. Intelligent Computers and Applications, 2021, 11(11).
[2] Du Yingkui, Zhang Yifang, Yuan Zhonghu, et al. Analysis of air pollution prediction accuracy of LSTM network by data Preprocessing [J]. Computer and Digital Engineering, 2021, 49(7).
[3] Zhang Xiaoqin, Cheng Yuying. Missing value filling method of component data based on random forest model [J]. Applied Probability and Statistics,2017,33 (1).
[4] Atiq R, Fariha F, Mahmud M, et al. A Comparison of Missing Value Imputation Techniques on Coupon Acceptance Prediction[J].
International Journal of Information Technology and Computer Science(IJITCS), 2022, 14(5).
[5] Zhang Mingwei, Zhang Tianyi, Zhong Ming, et al. The significance of arterial damage in the early detection of diabetes mellitus
verified by the integrated learning algorithm Stacking [J]. Chinese Journal of Medical Physics,2022,39(8).
[6] Shi Yuntao, Ren Peng, Li Shuqin, et al. Safety risk analysis and prediction of active meat and meat products based on Ensemble
learning [J]. Journal of Food Safety and Quality Inspection, 2019,13(16).
© 2025 by the EnPress Publisher, LLC. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.

Copyright © by EnPress Publisher. All rights reserved.

TOP