Discrete models for data imputation

摘要: The paper is concerned with the problem of automatic detection and correction inconsistent or out range data in a general process statistical collection. proposed approach able to deal hierarchical containing both qualitative quantitative values. As customary, erroneous records are detected by formulating set rules. Erroneous should then be corrected, modifying as less possible data, while causing minimum perturbation original frequency distributions data. Such called imputation. By encoding rules linear inequalities, we convert imputation problems into integer programming problems. procedure tested on real-world case census. Results extremely encouraging from computational quality point view.