Univariate economic time series forecasting by connectionist methods
This paper addresses the problem of time series forecasting by using a neural network. A predictor model able to extract the relevant regularities of univariate economic time series has been derived. Satisfactory performance comparison with a widely used statistical method confirms the feasibility of the connectionist approach.
Bibliographic Reference: Paper presented: INNC 90, Paris International Neural Network Conference, Paris (FR), July 9-13, 1990
Availability: Available from (1) as Paper EN 35498 ORA
Record Number: 199011246 / Last updated on: 1994-12-01
Original language: en
Available languages: en