Missing value simple imputation in matricial data using the singular value decomposition

Decision making in different areas of knowledge requires a correct application of statistical methodologies, however, in statistical data analysis there are quite common problems such as the presence of outliers and lack of information. The literature provides a large quantity of imputation methods, most of them with assumptions, but there are others that have no structural or distributional requirements like the Krzanowski method. This algorithm is based on the singular value decomposition and this project seeks for variants and iterative procedures that allow the reduction of uncertainty in the imputation process.