Abstract—Classification methods have been applied in real life problems. Real world dataset may contain missing values but many classification methods need complete datasets. Hence many missing value imputation methods like clustering based imputation method were proposed in literature. But when objects with missing values are high then the available complete objects in the dataset are less. Imputing missing values with limited amount of complete objects may not give good results when imputation is performed using complete objects only. The number of complete objects can be increased by treating the imputed object as complete object and using the imputed object for further imputations along with the available complete objects. In this paper we propose a missing value imputation method based on K-Means and nearest neighbors. This method uses the imputed objects for further imputations. The proposed method has been applied on clinical datasets from UCI Machine Learning Repository.
Index Terms—Missing value imputation, using imputed values, K-means, Nearest neighbors.
TThe authors are with the Department of Electronics and Computer Engineering, Indian Institute of Technology Roorkee, Roorkee, India (email: email@example.com).
Cite: Satish Gajawada and Durga Toshniwal, "Missing Value Imputation Method Based on Clustering and Nearest Neighbours," International Journal of Future Computer and Communication vol. 1, no. 2, pp. 206-208, 2012.