Aplicaciones de los modelos de predicción bio-inspirados en la administración

Abstract

Organizations and their environments are complex systems. Those systems are difficult to understand and to predict. Despite of this, the prediction is a key task for the enterprises’ management and for the decision making which always implies risk. The classical methods of prediction (among which are: the linear regression, Autoregresive Moving Average and exponential smoothing) establish assumptions such as linearity and stability for being mathematically and computationally tractable. By different means, however, have been shown the limitations of such methods. So well, in recent decades new prediction methods have emerged in order to encompass the complexity of organizational systems and their environments, instead of avoiding it. Among them, the most promising are the bio-inspired prediction methods (eg. neural networks, genetic / evolutionary algorithms and artificial immune systems). This article aims to establish a situational state of the actual and potential applications of bio-inspired prediction methods in management.