Missing values are a common issue in data collection that can significantly impact the performance of machine learning models. Addressing these missing values through various imputation methods is crucial for maintaining data integrity and model accuracy. This study evaluates the effectiveness of different imputation techniques, focusing on the Naive Bayes classifier for categorical data and comparing it with a baseline mode imputation approach. Additionally, the study examines mean and median imputations for numerical data. Using a chronic kidney disease dataset, we explore the performance of these imputation methods on decision tree and k$-$NN models. The results indicate that the decision tree model achieves higher accuracy with Naive Bayes imputation, whereas the k$-$NN model performs better with baseline imputation. Our findings suggest that the choice of imputation method should consider the specific classifier to optimize predictive performance. This research highlights the importance of tailored imputation strategies in enhancing the effectiveness of machine learning models dealing with missing data.