This paper describes the research and development of an artificial neural network system as a decision aid for employee selection. The ability of the artificial neural network to recognize patterns even using noisy data for employee selection and performance evaluation suggests this framework has significant potential advantage over traditional statistical models, such as regression analysis. Further, the neural model eliminates several methodological problems associated with the use of multiple regression, including non- linearity, incorrect function form specification, and heteroskedasticity.
Kirby, Edward; Dufner, Donna; and Palmer, Jonathan, "An Analysis of Applying Artificial Neural Networks for Employee Selection" (1998). AMCIS 1998 Proceedings. 27.