Uncertainty Aware Approach for Multiple Imputation Using Conventional and Machine Learning Models: A Real-World Data Study