Tourism Time Series Forecast - Different ANN Architectures with Time Index Input Academic Article Conference Paper uri icon


  • The main objective of this research is the identification of homogeneous groups within wind farms of a major operator playing in the energy sector in Portugal, based on two multivariate analyses: Hierarchical Cluster Analysis and Discriminant Analysis, by using two independent variables: annual liquid hours and net production. From the produced outputs there were identified three homogenous groups of wind farms: (1) medium Installed Capacity and Induction Generator based Technology, (2) high Installed Capacity and Synchronous Generator based Technology and (3) medium Installed Capacity and Synchronous Generator based Technology, which includes the wind farms with the higher annual liquid hours. It has been found that the results obtained by cluster analysis are well classified, with a total percentage of correct classification of 97,1%, which can be considered excellent.
  • Tourism demand is usually characterized by the time series of the “Monthly Number of Guest Nights in the Hotels”. Considering the increasing importance of this sector of activity, the prediction tools became even more relevant for public and private organizations management. Artificial Neural Networks (ANN) are a competitive model compared to other methodologies such the ARIMA time series models or linear models. In this paper the feedforward, cascade forward and recurrent architectures are compared. The input of the ANNs consists of the previous 12 months and two nodes used to the year and month. The three architectures produced a mean absolute percentage error between 4 and 6%, but the feedforward architecture behaved better considering validation and test sets, with 4,2% error.

publication date

  • January 1, 2012
  • January 1, 2013