Forecasting with Limited Information: ARIMA Models of the Trailer on Flatcar Transportation Market

Abstract

This paper evaluates the appropriateness of autoregressive integrated moving average (ARIMA) time-series models for forecasting in information-scarce environments. Such environments are defined and a specific transportation sector example examined. It is shown that ARIMA models adequately characterize economic activity in this example. AIUMA model forecasts are seen to dominate alternative models, and to provide information required for market efficiency. Specific attention is focused on a procedure for isolating potential model misspecification and on the need for post-sample forecast quality analysis.

A comparative Study of Forecasting the Electrical Demand in Basra city using Box-Jenkins and Modern Intelligent Techniques

“…The Fig. indicates that the electricity demand has high order nonlinearity and the time series has trends clearly, it turns to be high and that means the series is non-stationary, so the difference should be taken and the value of (d) should determine according to equations (19)(20)(21)(22)(23)(24) , and the series be stationary after the first difference (when d=1). As appears in Fig.…”

“…After testing the Box-Jenkins's symbols of the study data appear that the symbol which gave a close result to the original data style is Multiplicative Seasonal Model, and that because the study data contain both properties( trends, seasonality) furthermore, exist some random fluctuation [19]. This model uses for the time series have autocorrelation values non -zero,…”

“…xR i-3 R and each input variable was denoted with two membership MFs small and large. The parameters of the system are optimized by using ANN training, and structure is calculated by expert knowledge [2,19].…”

“…The time series can be defined as a variety of observations related with each other, and made sequentially in time to represent the manner of specific phenomena in past period of time, so if (YR t R ) is a symbol represents the series then [19]:…”

De Kunst Van Het Voorspellen:

“…De historische of niet-causale voorspelmethoden lenen zich vooral voor korte-termijnvoorspellingen wanneer veelal onvoldoende externe informatie voorhanden is om causale voorspellingen te maken. Toepassingen bieden Dent en Swanson (1978) voor de vervoersmarkt in de VS en Fase (1981a, b) voor resp. bankbiljetten en munten in Nederland.…”