An Expert System for Identification of Forecasting Model for Time Series

Document Type: Research Paper



Identifications and
analysis of time series are time consuming, based on trial and error and highly
dependent on expert judgments. This is mainly due to the presence of various
models for forecasting time series, as well as introducing new techniques for
analysis and predictions. In this paper, expert system structure is used to
replace traditional methods of model identifications for time series. Firstly,
several search engines are defined and analytical methods are specified. Next,
the knowledge base is developed such that a proper model can be assigned to
each data set. The goodness of fit is then evaluated by mathematical indices.
Repeating the process and modifying the responses to account for uncertain
situations, will provide a set of models to make the final decision. Lastly,
the performance of the proposed expert system is verified by a series of sample
data as a case study and the efficiency of the system is approved.