Comparison of Missing Data Imputation Methods in Time Series Forecasting

Time series forecasting has become an important aspect of data analysis and many real-world applications. However, undesirable missing values are often encountered, which may adversely affect tasks. In this study, we evaluate compare the effects imputation methods for estimating in a time series. Our approach does not include simulation to generate pseudo-missing data, but instead perform on actual measure performance model created therefrom. experiment, therefore, several models trained using different training datasets prepared each method. Subsequently, is evaluated by comparing accuracy models. The results obtained from total four experimental cases show that -nearest neighbor technique most effective reconstructing contributes positively compared with other methods.