Analysis of the granting of micro-credits to SMEs using the artificial neural network (ANN) method: the case of 29 SMEs of TREICHVILLE (Ivory coast)

The purpose of this article is to propose the advantage of using ANRs as a risk classification and prediction tool in the financing of SMEs in Ivory coast. The credit scoring is not optimal in terms of forecasting and classification "Steve Ambler & Alain Paquet (1997), Abdou et al (2007)" [1,2]. Indeed, the RNAs consist of a better classification-client for the granting of microcredit, in terms of their performance compared to traditional and traditional statistical methods "Marcus Odom & Ramesh Sharda (1990)" [3]. A two-stage survey (qualitative survey of the target population to identify important variables and collected by questionnaires and interview guides developed on the basis of selected variables) allowed us to select a sample of 29 tertiary sector firms. in the field of services with a margin of error of 5% and 2%, representing the rate of small firms in the tertiary sector holding normal accounts in Ivory Coast during the 2013 financial year.