-
Bayesian Model Averaging: An Application to the Determinants of Airport Departure Delay in Uganda
Wesonga Ronald,
Nabugoomu Fabian
Issue:
Volume 3, Issue 1, January 2014
Pages:
1-5
Received:
9 November 2013
Published:
10 December 2013
Abstract: Bayesian model averaging was employed to study the dynamics of aircraft departure delay based on airport operational data of aviation and meteorological parameters collected on daily basis for the period 2004 through 2008 in matrix X. Models were evaluated using the R programming language mainly to establish the combinations of variables that could formulate the best model through assessing their importance. Findings showed that out of the sixteen covariates, 62.5% were suitable for model inclusion to determine aircraft departure delay of which 40% exhibited negative coefficients. The following parameters were found to negatively affect departure delay; number of aircrafts that departed on time (-0.562), number of persons on board of the arriving aircrafts (-0.002), daily average visibility (-0.001) and year (-1.605). Comparison between Posterior Model Probabilities (PMP Exact) and that based on Markov Chain Monte Carlo (PMP MCMC) revealed a high correlation (0.998; p<0.01).The study recommended the MCMC as providing a more efficient approach to modelling the determinants of aircraft departure delay at an airport.
Abstract: Bayesian model averaging was employed to study the dynamics of aircraft departure delay based on airport operational data of aviation and meteorological parameters collected on daily basis for the period 2004 through 2008 in matrix X. Models were evaluated using the R programming language mainly to establish the combinations of variables that could...
Show More
-
Using GLS to Generate Forecasts in Regression Models with Auto-correlated Disturbances with simulation and Palestinian Market Index Data
Samir K. Safi,
Ehab A. Abu Saif
Issue:
Volume 3, Issue 1, January 2014
Pages:
6-17
Received:
12 December 2013
Published:
30 December 2013
Abstract: This paper involves an important statistical problem concerning forecasting in regression models in time series processes. It is well known that the most famous method of estimating and forecasting is the Ordinary Least Squares (OLS). OLS may be not the optimal in this context. So over the years many specialized estimation techniques have been developed, for example Generalized Least Squares (GLS). We are comparing the forecasting based on some estimators with the prediction using the GLS estimate. This comparison will be used by what is known as measures of forecast accuracy. We conduct an extensive computer simulation time series data, to make comparison among these methods. The similar forecasting criteria were developed and evaluated for the real data set on daily closing price in the Palestinian market index (Alquds Index). The data consists of 164 monthly observations and obtained from the website of the Palestine Stock Exchange. The main finding is that, for forecasting purposes there is not much gained in trying to identifying the exact order and form of the auto-correlated disturbances by using GLS estimation method. In addition, we noticed that the accuracy of forecasting using GLS method does not differ substantially than the other methods as Maximum Likelihood Estimation (MLE), Minimize Conditional Sum of Squares (CSS) and the combination of these two methods. Moreover, for parameter estimation, the GLS is nearly as efficient as the exact parameter estimation. On the other hand, the Ordinary Least Squares (OLS) method performs much less efficient than the other estimation methods and producing poor forecasting accuracy.
Abstract: This paper involves an important statistical problem concerning forecasting in regression models in time series processes. It is well known that the most famous method of estimating and forecasting is the Ordinary Least Squares (OLS). OLS may be not the optimal in this context. So over the years many specialized estimation techniques have been deve...
Show More
-
The G-Method for Rank Determination in Rank Order Statistics for ‘C’ Samples
Oyeka Ikewelugo Cyprian Anaene.,
Nwankwo Chike H.,
Awopeju K. Abidemi
Issue:
Volume 3, Issue 1, January 2014
Pages:
18-24
Received:
8 October 2013
Published:
10 January 2014
Abstract: This paper proposes a statistical method called ‘the G method’ to highlight its generalized nature for the determination and assignment of ranks to sample observations drawn from several populations for possible use in further analyses. The sampled populations may be measurements on as low as the ordinal scale and need not be continuous or even numeric. The proposed rank determination statistical model intrinsically and structurally provides for the breaking of possible ties between sample observations and automatically assigning such observations their mean ranks. This approach and hence the proposed model therefore obviate the need for the sampled populations to be continuous. They may be discrete or even non-numeric measurements on as low as the ordinal scale. The proposed method is of more generalized and wider applicability than an existing formulation which can be used with only continuous populations and is easier to use in practice than the usual traditional method which is often tedious and cumbersome, especially with large samples. The proposed method is illustrated with some data and shown to yield the same results as other existing methods where these methods are equally applicable.
Abstract: This paper proposes a statistical method called ‘the G method’ to highlight its generalized nature for the determination and assignment of ranks to sample observations drawn from several populations for possible use in further analyses. The sampled populations may be measurements on as low as the ordinal scale and need not be continuous or even num...
Show More
-
Bayesian Estimation of Reliability Function for A Changing Exponential Family Model under Different Loss Functions
P. Nasiri,
N. Jafari,
A. Jafari
Issue:
Volume 3, Issue 1, January 2014
Pages:
25-30
Received:
30 November 2013
Published:
28 February 2014
Abstract: The paper deals with estimating shift point which occurs in any sequences of independent observations x1, x2, …, xm, xm+1, …, xn of poisson and geometric distributions. This shift point occurs in the sequence when xm i. e. m life data are observed. With known shift point 'm', the Bayes estimator on befor and after shift process means θ1 and θ2 are derived for symmetric and assymetric loss functions. The sensitivity analysis of Bayes estimators are carried out by simulation and numerical comparisons with R-programming. The results show the effectiveness of shift in sequences of both poisson and geometric distributions.
Abstract: The paper deals with estimating shift point which occurs in any sequences of independent observations x1, x2, …, xm, xm+1, …, xn of poisson and geometric distributions. This shift point occurs in the sequence when xm i. e. m life data are observed. With known shift point 'm', the Bayes estimator on befor and after shift process means θ1 and θ2 are ...
Show More
