An Intelligent Missing Data Imputation Techniques: A Review

The incomplete dataset is an unescapable problem in data preprocessing that primarily machine learning algorithms could not employ to train the model. Various imputation approaches were proposed and challenged each other resolve this problem. These imputations established predict most appropriate value using different with various concepts. Furthermore, accurate estimation of method exceptionally critical for some datasets complete missing value, especially imputing medical data. purpose paper express power distinguished state-of-the-art benchmarks, which have included K-nearest Neighbors Imputation (KNNImputer) method, Bayesian Principal Component Analysis (BPCA) Multiple by Center Equation (MICE) denoising autoencoder neural network (MIDAS) method. methods contributed achievable resolution optimize evaluate points value. We demonstrate experiment all these techniques based on same four are collected from hospital. Both Mean Absolute Error (MAE) Root Square (RMSE) utilized measure outcome implementation compare prove extremely robust overcomes problems. As a result experiment, KNNImputer MICE performed better than BPCA MIDAS imputation, has algorithm.