Comprehensive Study of Data Imputation Techniques For Machine Learning Models

Missing data is a significant challenge in machine learning, particularly in the development of propensity models where accurate predictions depend on complete and reliable data. This paper provides a comprehensive exploration of various imputation techniques tailored for machine learning workflows, specifically in the context of propensity modeling. Each technique is categorized by its applicability to different types of data and scenarios of missingness. The goal is to equip practitioners with the tools and knowledge to effectively handle missing data, ensuring robust and accurate propensity models