Combined application of the hierarchical decision process with time series analysis: a telecommunications industry forecasting application

Summary form only given, as follows. The author incorporates elements of two types of forecasting processes, time series analysis and the hierarchical decision process (HDP), to optimize the advantages of each forecasting telephone lines and stations in use for the Anchorage Telephone Utility (ATU). A particular service area was selected because results of a time series forecasting model revealed a declining service requirement for this area, which was contrary to the opinions of several experienced ATU forecasters and managers. An expert panel was chosen. The HDP model for evaluating decisions was applied by the panel to determine the relative influence of several changes in the telecommunications market on the future growth of a specific service area. Relative growth rates generated from the HDP model were combined with results of a time series analysis to produce a new service requirement forecast. Results indicate that the combination of historical, object data and subjective, judgmental factors can be useful in projecting quantities with multiple influencing factors. The process is beneficial both in terms of increased accuracy of the forecast and in terms of providing an improved management decisionmaking tool