Regresar

Algorithm for the selection of attributes in the pbkp_rediction of business failure

Abstract:

The selection of relevant attributes is a common task in data mining. In pbkp_redicting business failure it is useful to have techniques that allow you to select financial reasons that have high pbkp_redictive power. In this paper an algorithm is proposed that in each iteration selects a subset of attributes through random matrix decomposition. Each subset of attributes is evaluated by its pbkp_redictive capacity calculated using the technique k closest neighbors with cross validation. The performance of the algorithm is tested with a set of data that contains financial reasons for failed companies and non-failed companies. The results obtained are comparable with those reported in the literature for the same data set.