Missing Value Imputation Using Adaptive Fuzzy K-Nearest Neighbour based Support Vector Machine in Healthcare

Healthcare professionals play a vital role in diagnosing and treating a wide range of medical conditions, including injuries, diseases, and both physical and mental impairments.Their expertise not only assists in the early detection of health problems but also ensures that patients receive personalized care.However, missing values in healthcare data result in incomplete or biased analysis, affecting the accuracy of patient diagnosis, research outcomes, and treatment planning.This research proposes an Adaptive Fuzzy K-Nearest Neighbor-based Support Vector Machine (AFKNN-SVM) for missing value imputation in healthcare.In KNN, the adaptive fuzzy model is incorporated to handle missing data by assigning flexible degrees to neighbors, which helps in managing uncertain or overlapping data.SVM is used to classify the data, minimizing bias and improving the model's ability to generalize from training data.The proposed AFKNN-SVM achieves improved accuracies of 99.56%, 93.23%, and 97.27% on the Pima Indians Diabetes Database, Mammographic, and Dermatology datasets, respectively, compared to existing methods such as KNN with the Tri-ensemble voting model.