Time Series Analysis, Forecasting and Control
Time Series Analysis, Forecasting and Control
Computer Algorithms
Time series applied in Romanian economy
AIKED'09 Proceedings of the 8th WSEAS international conference on Artificial intelligence, knowledge engineering and data bases
Hi-index | 0.00 |
Understanding the structure of a Temporal Series is essential for the Finance Engineer or the manager of a company in order to make a correct decision. In addition to the insights of the time series structure, the foundations of Temporal Series Theory may help to predict the future using statistics inference. The outstanding Integrated Auto Regressive Moving Average Model ARIMA(p, d, q) is widespread and very used in Finance and Economics. Nevertheless, the process of determining its p, d and q parameters is done manually and prone to error. This paper proposes an algorithm, developed in the R statistical package, which tests all the possibilities, defined by the analyst, for a Multiplicative Seasonal ARIMA Model (SARIMA (p, d, q) x (P, D, Q)). The algorithm sorts the best alternatives considering different objective functions, like Akike Information (AIC), log likelihood function criteria or the alternatives that presents the lowest predicted quadratic errors.