Hiwa Farughi - Department of Industrial Engineering, University of Kurdistan Sanandaj, Iran
Ahmad Hakimi - Department of Industrial Engineering, University of Kurdistan Sanandaj, Iran

In this paper, one of the most important criterion in publicservices quality named availability is evaluated by usingartificial neural network (ANN). In addition, the availabilityvalues are predicted for future periods by using exponentialweighted moving average (EWMA) scheme. Results basedon comparative studies between four methods based on ANNand by considering the several conditions for the effectiveparameters in ANN show that, the generalized regressionmethod is the best method for predicting the availability.Furthermore, the results of the EWMA method is also showsthe suitable performance for predicting the availability valuesin future periods.

Maintenance, availability, prediction, artificial neural network, exponentially weighted moving average