Design classification under missingness and imbalance

You are tasked with developing a model that predicts whether a patient will be readmitted to the hospital within 30 days, using information extracted from claims and EHR data. Approximately 30% of the values are missing; the absence may be concentrated in specific columns, specific rows, or both. The positive class is uncommon, making up about 6% of cases. Propose a detailed approach that handles each of the following, including justifications and trade-offs: Missing data.…