Comparison of Missing Data Imputation Methods in Time Series Forecasting |
| |
Authors: | Hyun Ahn Kyunghee Sun Kwanghoon Pio Kim |
| |
Affiliation: | 1.College of Civil Engineering, Changsha University of Science and Technology, Changsha, 410114, China2 College of Civil Engineering, Chongqing University, Chongqing, 730000, China3 Department of Civil and Environmental Engineering, University of Auckland, Auckland, 1010, New Zealand4 Hunan Construction Engineering Group Co. Ltd., Changsha, 410004, China |
| |
Abstract: | Time series forecasting has become an important aspect of data analysis and has many real-world applications. However, undesirable missing values are often encountered, which may adversely affect many forecasting tasks. In this study, we evaluate and compare the effects of imputation methods for estimating missing values in a time series. Our approach does not include a simulation to generate pseudo-missing data, but instead perform imputation on actual missing data and measure the performance of the forecasting model created therefrom. In an experiment, therefore, several time series forecasting models are trained using different training datasets prepared using each imputation method. Subsequently, the performance of the imputation methods is evaluated by comparing the accuracy of the forecasting models. The results obtained from a total of four experimental cases show that the -nearest neighbor technique is the most effective in reconstructing missing data and contributes positively to time series forecasting compared with other imputation methods. |
| |
Keywords: | Missing data imputation method time series forecasting LSTM |
|
| 点击此处可从《》浏览原始摘要信息 |
|
点击此处可从《》下载全文 |
|