On Predicting Internal Humidity Missing in Emergency Rooms Using Environmental Data from Korea Meteorological Administration

Bumsik Lee, Gwangrae Yeom, Jongpil Jeong

Abstract: This study addresses the issue of missing environmental data estimates collected from sensors within ER (Emergency Room). Ignoring missing in data during analysis can introduce bias and lead to incorrect conclusions. There are various methods to tackle this problem, such as deletion, statistical imputation, machine learning-based imputation, and generative imputation. In particular, compensating for missing data requires a multi-pronged approach to predicting methods. In this paper by combining external data from the Korea Meteorological Administration with internal ER’s internal environmental data, we estimated the missing variable of internal humidity. It integrates external data with the internal data to analyze correlations and predict missing values. The analysis revealed a strong correlation between internal humidity and external temperature. There are two types of missing values: short-term and long-term. We focused on addressing long-term missing values using machine learning. The outcomes of this paper can serve as a valuable resource for enhancing safety and the overall environment within emergency departments. Consequently, this research is anticipated to enhance the emergency department’s environment, ultimately contributing to the safety and comfort of patients in critical situations. Additionally, it can furnish ER operators with vital information for decision-making, and afford patients an improved experience.