Context:

A set of options is proposed to help determine the most appropriate method to correct in databases of appreciable size, initial conditions of missing data and that will be used in research processes.

Methodology:

This article addresses a proposal for the development and management of robust databases such as financial records, focusing from the Knowledge Discovery in Data bases (KDD) process.

Results:

A methodology is developed and tested using three imputation techniques in a database built from 1,253,280 financial records of 2,238 companies that represent seven years of their economic activity in the town of Chapinero in the city of Bogotá D.C.

Conclusions:

A comparison of the imputation methods is carried out as a determining factor for the choice of the imputation method and consolidation of the base for later use.

Financing:

Fundación universitaria Los Libertadores