An artificial neural network (ANN) is made up of neurons called usually processing unit, cells or nodes. An artificial neuron receives signals at its inputs (the dendrites of biological neuron), each of them having assigned a weight (for the biological neuron this signals are in fact electrical pulses, the strength of the signal correspond to the weight of the artificial neuron), computes a weighted sum of these signals and the neuron fires if the result exceeds a value called activity level, propagating the signal toward the output (the axon of the biological neuron).[1] ANN offer qualitative methods for business and economic systems that traditional quantitative tools in statistics and econometrics cannot quantify due to the complexity in translating the systems into precise mathematical functions. Training of an ANN is a laborious process that stops once a performance criterion is satisfied. If an ANN is over trained, a curve-fitting problem may occur whereby the ANN starts to fit itself to the training set instead of creating a generalized model. This typically results in poor predictions of the test and validation data set. On the other hand, if the ANN is not trained for long enough, it may settle at a local minimum, rather than the global minimum solution. This typically generates a suboptimal model. [2], [3] NeuroShell offers to the users two main interfaces: the beginner’s interface and the advanced interface. The advanced neural network interface offers the possibility to choose among the ANN architectures available. These architectures are grouped in two main categories: classification nets and predictive nets. The simulation process begins with the data set that can be imported from other programs (Excel for example) or filled using the module special designed. The problem having its data set, there must be established the number of inputs and outputs, a test data must be extracted in order to save the best results of the learning process to this test data set. The results of the learning are applied using the neural network which numbers of inputs, outputs, hidden neurons, weights are already set, and the results of applying the neural network for the problem are visualized using the Examine Data module. [4] The results obtain after training the network on a test data set is shown below.figura