The interaction between imputation and regression models

A significant problem and source of great discomfort throughout the data analysis process might come from missing values in the dataset. One of the areas where missing data is most common is the medical field. In earlier works, the proposed approach removes the missing value in a list-wise [1] or pairwise [2] available recording. Another commonly used method in missing data analysis is the imputation method, which replaces the missing data with alternative values [2,3]. The performance of various regression models on missing data has been compared. Their performance will be evaluated on imputed and non-imputed data combinations, and the most appropriate imputer+regressor or no-imputer+regressor pairs will be determined.
To summarize, in this study, we will seek answers to the following questions:
• How can we achieve a more accurate regression model in the presence of missing data with imputation or without?
1. Schafer, J.L.: Analysis of incomplete multivariate data. CRC press (1997).
2. van Buuren, S.: Flexible imputation of missing data. Chapman and Hall/CRC, edn. (2018).
3. Emmanuel, T., Maupong, T., Mpoeleng, D., Semong, T., Mphago, B., Tabona, O.: A survey on missing data in machine learning. Journal of Big Data 8(1), 1–37 (2021).