This study examined the role of ICT Education in Harnessing Human Capital Development (HCD) in Africa. It considered the vital component of ICT as it relates to Education in a vision towards an intelligent knowledge based society. ICT as an engine for promoting HCD, financial stability, enhancing wealth creation, job opportunities, and management of epidemic diseases such as Corona virus (Covid-19), Malaria, Hepatitis B, HIV/AIDS, Ebola Virus. Developed ICT policies and strategies are driven skills that harness productivity as well as impact the development of an informed society. However, inadequate knowledge, failure to implements policies, corruption, in-adequate man power work-force in education, are some of the critical challenges that have placed Africa far below average in rating. And proposed a model on the role of ICT education as way forward for harnessing HCD in Africa. The result showed that improvements on ICT Education and HCD significantly enhance Africa’s potentials in Science, Technology, and Humanities for Sustainable Development.This study examined the role of ICT Education in Harnessing Human Capital Development (HCD) in Africa. It considered the vital component of ICT as it relates to Education in a vision towards an intelligent knowledge based society. ICT as an engine for promoting HCD, financial stability, enhancing wealth creation, job opportunities, and management of epidemic diseases such as Corona virus (Covid-19), Malaria, Hepatitis B, HIV/AIDS, Ebola Virus. Developed ICT policies and strategies are driven skills that harness productivity as well as impact the development of an informed society. However, inadequate knowledge, failure to implements policies, corruption, in-adequate man power work-force in education, are some of the critical challenges that have placed Africa far below average in rating. And proposed a model on the role of ICT education as way forward for harnessing HCD in Africa. The result showed that improvements on ICT Education and HCD significantly enhance Africa’s potentials in Science, Technology, and Humanities for Sustainable Development.

Africa, Development, Education, HCD and ICT

This study used a total of 161 countries which were stratified into six regions comprising Africa, Asia, Europe, North America, South America, and Australia/Oceania. We critically described the statistical view on the COVID-19 coronavirus epidemic across the world by giving the facts and figures on the number of deaths, percentage of deaths to COVID-19, infected cases, percentage infected to population and the recovered cases in the world as at Saturday, 21 March 2020. Algeria has the highest number of deaths while Mauritius has the highest percentage of death and Egypt has the highest number of infected individuals as well as the highest number of recovered cases. In Asia, China has the highest number of deaths and individuals infected while Iran has the highest percentage of death and percentage of individuals infected with COVID-19. The maximum recovered cases were 71,857 representing the recovered cases in China in Asia. Italy has the highest number of deaths and the number of infected individuals across Europe to COVID-19 while San Marino has the highest percentage of death and percentage of the population infected with COVID-19. The highest recovered cases were 4,440 in Italy. In North America, the United States has 275 deaths as the highest number of deaths to COVID-19 and also has the highest percentage deaths to population, the highest number of infected individuals across North America and also has the percentage of the population infected with COVID-19 to be 0.006%. The highest recovered cases were 147 in the United. Brazil with 15 deaths has the highest number of deaths while Guyana has the highest percentage of death to COVID-19 in South America. The country with the highest number of infected individuals across South America was Brazil while Uruguay has the percentage of the population infected with COVID-19 to be 0.003%. Chile has four recovered cases which were the highest cases recorded at that time in South America. The country with the highest number of infected individuals across Australia/Oceania region was Australia while it also had the highest percentage of the population infected with COVID-19. The recovered cases were only recorded as two in Australia.

COVID-19, Epidemic, Deaths, Percentage, Recovered cases

This study used a total of 161 countries which were stratified into six regions comprising Africa, Asia, Europe, North America, South America, and Australia/Oceania. We critically described the statistical view on the COVID-19 coronavirus epidemic across the world by giving the facts and figures on the number of deaths, percentage of deaths to COVID-19, infected cases, percentage infected to population and the recovered cases in the world as at Saturday, 21 March 2020. Algeria has the highest number of deaths while Mauritius has the highest percentage of death and Egypt has the highest number of infected individuals as well as the highest number of recovered cases. In Asia, China has the highest number of deaths and individuals infected while Iran has the highest percentage of death and percentage of individuals infected with COVID-19. The maximum recovered cases were 71,857 representing the recovered cases in China in Asia. Italy has the highest number of deaths and the number of infected individuals across Europe to COVID-19 while San Marino has the highest percentage of death and percentage of the population infected with COVID-19. The highest recovered cases were 4,440 in Italy. In North America, the United States has 275 deaths as the highest number of deaths to COVID-19 and also has the highest percentage deaths to population, the highest number of infected individuals across North America and also has the percentage of the population infected with COVID-19 to be 0.006%. The highest recovered cases were 147 in the United. Brazil with 15 deaths has the highest number of deaths while Guyana has the highest percentage of death to COVID-19 in South America. The country with the highest number of infected individuals across South America was Brazil while Uruguay has the percentage of the population infected with COVID-19 to be 0.003%. Chile has four recovered cases which were the highest cases recorded at that time in South America. The country with the highest number of infected individuals across Australia/Oceania region was Australia while it also had the highest percentage of the population infected with COVID-19. The recovered cases were only recorded as two in Australia.

COVID-19, Epidemic, Deaths, Percentage, Recovered casesCOVID-19, Epidemic, Deaths, Percentage, Recovered cases

This research “simplified chi – square (C- square)” is carried out to remove the difficulties in the conventional chi-square. In this paper, a chi-square statistic is simplified to obtain a statistic called C^2. C-square is the sum of square of the co-factor divided by marginal totals multiply by ground total of any given categorical data in which the estimate obtained from the statistic is almost the same with the estimate from chi-square. C^2or C-square can be derived from chi-square and also be obtained through co-factor method This research work is carry out in order to check the cumbersome and difficulties in chi-square manual computation and minimize the time spending in it computation and also reduce the work load, data entering and time taking for R-programming to execute the assignment. When C^2or C-square is used with several examples we discovered that it is faster and the result is approximately the same as the result obtain in R programming and conventional chi – square.

C-square, chi-square, cofactor, statistic, R-programming

The exclusion of inmates from the use of ICTs and the internet can only continue to be counterproductive and promotes the cyclic return of discharged inmates. 394 National Open University of Nigeria inmate students are excluded from the benefits of online learning due to the security protocols as a result of the policy on the prohibition of internet usage by inmates. Nigerian Correctional Service data between 2013-2017 was employed; quantitative methods were employed for the study. The study revealed that men were significantly more likely to recidivate than women (48% versus 19%). Also that out of the 394 registered inmate students in NOUN, 94.7% had only Ordinary level (O’level) as their highest qualifications and only 17 had graduated. The result alluded to the fact that increased engagement through online learning would not only result to improved enrollment and performance, it would equally reduce recidivist crimes in Nigerian Correctional Centres.

Open and Distance Learning, inmates, recidivism, Nigerian Correctional Service

The study of numerical solutions for nonlinear equations is one of the trending areas of research in pure and applied mathematics. As it is applied to several areas such as optimization, mathematical modelling, operation research and so on. This research work compared three methods (Bijection, Secant and Newton-Raphson) for solving nonlinear equations. It was observed that secant may converge faster than Bisection method, while Newton-Raphson method converge faster than secant method. Therefore, we concluded that Newton-Raphson method is the most efficient and fastest method for solving nonlinear equation than secant method and Bijection method because it converges at the three alterations.

Nonlinear, Bijection, Secant, Newton-Raphson, convergent

This article aims at depicting the genuity and usefulness of the amalgamated technique known as Homotopy perturbation with Mahgoub Transformation in unveiling the particular and coarse-grained solutions of the equation of Korteweg-de Vries of third order and nonlinear. Considered for illustrations were three cases with the proposed mechanism engaged to gain their coherent calculations. Acquired outcomes via the proposed mechanism were contrasted with solutions achieved by means of Adomian Polynomial with Elzaki Transformation and Elzaki transformation with Homotopy perturbation and were viewed as the equivalence of one another. Considering the snappy gathering of the results experienced by the method for the representative cases, it is therefore suggested that this mechanism is progressively effective, accommodating, productive and of high precision with lesser counts. Besides, it is of basic to acquaint the actuality that this article had incorporated another modus operandi to the existing means of disentangling equations of this kind.

Korteweg-de vries equation, Mahgoub transformation, Homotopy perturbation, nonlinear, partial differential equation

Phenomenon whose occurrences are described with count data in a multi-dimensional contingency table is common in medical and epidemiological studies. This study thus obtained the density function(s) of nonlinear vectors via mutual information approach and its measure of dependence from multivariate exponential power distribution which is a member of the multivariate elliptical contoured family. The obtained model which accommodates light and heavy-tailed member distributions depending on the shape parameter was used to establish results for nonlinear vectors from multivariate Laplace, normal, uniform, students’ t, and Fisher distributions. The Multivariate dependency of live birth on maternal age as the number of pregnancies increases irrespective of maternal height advantage as claimed from the previous study was established from the theoretical model obtained.

Multivariate elliptical contoured distribution, exponential power distribution, mutual information, nonlinear density model, entropy, live birth, maternal height

This project is focused on the study of some quality characteristics of robust tissue paper a product of Epesok Paper Mills using statistical quality control techniques. Three important quality characteristics are Ph, Basic weight (g⁄m^2 ) and Absorbency (mm) which are all variable measures. The data used for the study is a secondary data collected from the daily laboratory records of Epesok paper Mills Company of Nigeria for a period of seven (7) months. Control chart was used to monitor the performance of each of the quality characteristics. Also correlation analysis was carried out to determine possible correlation been pairs of the quality characteristics. The results shows that each of the characteristics monitored were statistically under control. Also the results of the correlation analysis confirm significant correlations among some quality characteristics.

Correlation, Control chart

This study examined the trends in the prevalence of breast cancer in Ilorin, Kwara State, Nigeria. Breast cancer is an uncontrolled growth of breast cells, Cancer occurs as a result of mutations, or abnormal changes, in the genes responsible for regulating the growth of cells and keeping them healthy. The focus of this research work was to evaluate the factors affect analyze some factors responsible for the survival of Breast cancer patient in Nigeria. Furthermore, this study used a total number of two hundred and six (206) dataset collected from the records of women between the ages of ≤15 and >45 years in the University of Ilorin Teaching Hospital (UITH), Kwara State. The methodology employed are non-parametric Kaplan-Meier method as well as Log-Rank statistic and the semi-parametric such as Cox proportional hazards regression model to analyze the data in order to determine the hazard ratios of the various covariates. The result obtained by the PHreg procedure indicated that stages of breast cancer, year of admission, and age of patients as factors influencing the outcome of breast cancer patient in Nigeria.

Survival Analysis, Breast Cancer, Kaplan-Meier Estimator, Log-Rank statistic, Cox proportional Regression, Nigeria

Heteroscedasticity occurs mostly because of beneath mistakes in variable, incorrect data transformation, incorrect functional form, omission of important variables, non-detailed model, outliers and skewness in the distribution of one or more independent variables in the model. All analysis were carried out in R statistical package using Imtest, zoo and package.base. Five heteroscedasticity tests were selected, which are Park test, Glejser test, Breusch-Pagan test, White test and Goldfeld test, and used on simulated datasets ranging from 20,30,40,50,60,70,80,90 and 100 sample sizes respectively at different level of heteroscedasticity (that is at low level when sigma = 0.5, mild level when sigma = 1.0 and high level when sigma = 2.0). Also, the significance criterion alpha = 0.05. However, each test was repeated 1000 times and the percentage of rejection was computed over 1000 trials. For Glejser test, the average empirical type I error rate for the test reject more than expected while Goldfeld has the least power value. Therefore, Glejser test has the highest capacity to detect heteroscedasticity in most especially on simulated datasets.

Park test, Glejser test, Breusch-Pagan test, White test, Goldfeld test

Coronavirus popularly known as COVID-19 has created a large scale of infections in Nigeria. To study the efficacy of drug response of COVID-19, we developed a continuous time stochastic model for the dynamics of infections assuming that the exposed individuals develop an active case within fourteen days. To generate the sample paths describing the Susceptible-Exposed-Infected-Recovered individuals, we use stochastic simulation using tau-leap simulator. To evaluate the efficacy of response drug to recover in treating the incidence of COVID-19 in Nigeria, we compared the predicted sample paths of coronavirus with and without ERDR. At 90 days administration, it reduced the prevalence to 85% recovery from an active exposed. To assess our model, we compared the tau-leap simulator with the coronavirus SEIR fit.

Coronavirus, drug efficacy, stochastic simulation, tau-leap simulator

Rapid growth of technology and prevalent use of computer in the business and other areas appeals for technology advancement, most organizations are converting their paper documents into electronic documents that can be processed by computer. Much research has been done on the identification of English, Arabic, Japanese and Chinese. It is observed that the research on recognition of printed characters for most of the African languages including Hausa is still an open research problem. However, Hausa language is one of the dominant languages in sub-Sahara Africa hence the study. Here, the study considers the processing of gray level images only, since they contain enough information to perform feature extraction and image analysis. Therefore, the paper develops a simple and efficient method for the recognition of isolated printed Hausa characters. Images of printed Hausa characters were captured using a scanner and images were pre-processed to remove noise. The methodology for isolated printed Hausa character recognition was based on efficient feature extraction followed by a suitable feature vector dimensionality reduction scheme. HMM algorithm was adopted to develop a system for recognition of printed Hausa characters.

DTW, Hausa Language, HMM, Speech Recognition, MFCC

This paper focuses on the investigation of traditional digital image watermarking with crt and zero watermarking with crt in the dct domain. Functional analysis of imperceptibility, robustness and time of execution is performed. Based on the result of the experiment the crt with zero watermarking provides better performance since there are no distortion with a faster time of execution, however, it is still susceptible to cropping attacks. Traditional crt watermarking still has a high tamper assessment function (taf) without any attack.

Chinese Remainder Theorem, Discrete Cosine Transform, Digital Watermarking

In this paper, we used the New Iterative Method (NIM) developed by Daftardar-Gejji and Jafari for the solution of linear and nonlinear systems of partial differential equations. This method is very simple as it reduces the size of computation and readily converges to the exact solution. To demonstrate the efficiency of the method, some illustrative examples were provided. The results obtained confirmed that the method is an efficient method for a wide variety of systems of linear and nonlinear PDEs.

New Iterative Method, Systems, Partial Differential Equations, Linear, Nonlinear

In this paper, a graphical solution of the nonlinear programming problem was presented the result satisfies the graphical solution of the nonlinear programming problem and we also solved some NLPP question in this paper by graphical method.

Non linear programming problem, Objective function, Constraint, Graphical solution, Feasible region

Designs that with the ability to remove heterogeneity in the experimental material, have received considerable attention in the literature. By convention, the parameters of row-column designs constructed in this paper are v, b, p, q which respectively denote the number of treatments, rectangular blocks, rows and columns; and indeed, each of the rectangular blocks has nested within it, p rows and q columns. Designs of series I and II which are block designs are re-configured in a systematic manner in this paper to give block designs with the ability to remove heterogeneity in multiple directions, namely row and column; and these row and column are also nested within each of available rectangular blocks. All the designs constructed are balanced with respect to two of the three components of the resulting designs namely, rectangular block and, column. This is consequently reflected in the concurrence of treatment pairs that were obtained for the row (ʎijr), column (ʎijc), and rectangular block (ʎijb) components of the designs respectively. Although, there were cases when some pairs of treatments do not occur together in the same block, row and column simultaneously, this does not violate connectedness property for all the designs constructed. Row-column designs constructed in this paper are for even and odd number of treatments, that is, v = 6, 7, 8, 9, 10, and 11. All the designs constructed are less restrictive with the attendant reduction in the required number of experimental units.

Nested design, Heterogeneity, Row and Column Blocking Variables, Concurrence of Treatment Pairs

This paper reviews the relationship between stock return and comprehensive income from published papers. Eight journals of accounting economics were randomly selected between a specific periods of (1981-2014). The objective of this study is to find out how consistencies are the researchers findings on stock return and comprehensive income through their relationship. Each of the paper was critically scrutinized with emphasis on Title, Method and Finding. The modality of finding the outcome was based on three hypotheses in which stock price as a function was enumerated. The result indicated that market development serve as prerequisite change to stock exchange. Thus the result as regard market is consistent.

Stock return, Comprehensive income, Stock prize, Market development, Market efficiency

A shared memory multi-core computer system is a computing paradigm in which processors have more than one core to process requests and also have access to a common memory. Most existing works are limited to modeling of a shared memory single-core computer system and thus the models are not flexible enough to study the operations of multi-core computer systems. Hence, in this paper, a high-level Petri Nets formalism (Timed Coloured Petri Nets) was used to develop a simulation model for a shared memory multi-core computer system. Intel HP core i5 was used as a case study in developing the TCPN model for a shared memory multi-core computer system. The developed TCPN model was simulated using Coloured Petri Net (CPN) tools. One hundred and fifty simulation runs were carried out in order to obtain average utilization rate of the shared memory and average waiting time of the processor’s cores in accessing the shared memory. The developed TCPN model was validated based on both real and simulated average memory utilization of the shared memory multi-core computer system. The validation result of the developed TCPN model showed that there was no significant difference between the simulated and real average memory utilization of the shared memory multi-core computer system.

Modelling, multi-core, petri net, shared memory, simulation

Violation of constant error variance assumption in any linear regression model leads to the problem of heteroscedasticity. In practices, variance of error terms are unequal and unknown in nature, but there is need to determine the presence or absence of this problem that do exist in unknown error term as a preliminary diagnosis on the set of data we are to analyze or perform hypothesis testing on. Although, there are several forms of heteroscedasticity and several detection methods of heteroscedasticity, but for any researcher to arrive at a reasonable and correct decision, best and consistent performed methods of heteroscedasticity detection under any forms or structured of heteroscedasticity must be determined. This paper consider seven heteroscedasticity structures that were originally proposed by different authors for developing statistical tools for heteroscedasticity detection in linear regression model. Nine heteroscedasticity detection methods were considered in other to determine some heteroscedasticity methods that are best to be used for the determination of unequal variance presence among the error terms in a linear regression model when there exist no forms of correlation between the exogenous variables of the model via the power of the test. In this work, Monte Carlo experiment was conducted one thousands (1000) times on a linear regression model with three predictor variable that exhibit no degree of multicollinearity (ρ=0) and seven sample sizes (n=15,20,30,50,100 and 250). The parameters of the model were specified to be β_0=4, β_1=0.4, β_2=1.5, β_3=3.6 and the various tests were examined at 0.1, 0.05 and 0.01 levels of significance. The Confident interval criterion (C.I) was used to determine the performances of the methods. The study concluded that when power of a test is considered to determine the preferred methods of heteroscedasticity detection in a model where there exist no multicollinearity between the exogenous variables at significance level of α = 0.1, BG test or GFQ test are the preferred methods, when testing under significance level of α = 0.05,the preferred method to used is GFQ method, while at α = 0.01, the preferred method to use in testing for the presence of heteroscedasticity is either BG method or NVST method.

Regression model, heteroscedasticity, heteroscedasticity detection methods, heteroscedasticity structures, significance levels, Confidence Interval and Power rates

In this paper, Magnetohydrodynamic (MHD) fluid flow in a porous channel under optically thick limit radiation was examined. The non-dimensional equations are solved constructing a semi-analytical numerical technique by differential transform method (DTM). DTM is useful in this work as it can be used to accurately solve larger class of linear and nonlinear problems. The effect of velocity and temperature for different values of physical parameter such as thermal Grashof number, radiation parameter, Magnetic parameter and porosity parameter is considered. It is observed that the velocity increases with increasing values of magnetic and decreases with increasing values of Grashof number and porosity parameter. Also, the temperature decreases with increasing values of magnetic and radiation parameter but decreases with Grashof number and porosity parameter. Furthermore, there was an increase in heat transfer due to thermal conduction as the radiation increases.

Magnetohydrodynamics, Heat Transfer, Radiation, Optically Thick Limit, Differential Transform Method

This work presents the partial least squares (PLS) regression model as applicable to a wide range of fields especially where there is a number of correlated explanatory variables. The strategy is to decompose variables into components. This can be contrasted with other methods such as principal component regression, ridge regression where vital variable may be accidentally deleted. It is also not dependent on any unknown factor upon formulating a model of two dependent variable and four independent variables. The model is subsequently applied to central bank of Nigeria data. The R^2 value of 0.9608 and 0.8607 were obtained for external reserves and surplus respectively. The work shows that partial least squares regression is a more robust and creditable model for multivariate regression model.

Partial least Squares, Responses, Predictor External reserve, Surplus

Statistical models of biased sampling of two non central hypergeometric distributions Wallenius' and Fisher's distribution. However, not many of the logic of hypergeometric distribution have been investigated by different techniques. The work examined the procedure of the two non-central hypergeometric distributions and investigates the statistical properties of the two non central hypergeometrics which includes the mean and variance that were obtained. The parameters of the distribution were used using the direct inversion method of hyper simulation of biased urn model in the environment of R statistical software with varying odd ratios (w) and group sizes (mi) were used. It was found that the two non-central hypergeometrics are approximately equal in mean, variance and coefficient of variation and differ as odds ratios (w) becomes higher and differfrom the central hypergeometric distribution with ω = 1. Furthermore, in univariate situation we observed that Fisher distribution at (ω = 0.2, 0.5,0.7, 0.9) is more consistent than Wallenius distribution although central hypergeometric is more consistent than any of them also in multinomial situation, it was observed that Fisher distribution is more consistent at (ω = 0.2, 0.5) Wallenius distribution at (ω = 0.7, 0.9) central hypergeometric at (ω = 0.2)

Non central hypergeometric, wallenius distribution, fisher distribution, univarate situation, multinomial situation

Credit scoring has become the source of many customers intake to the banking industry, due to the large intake of customers and huge database in the organisation credit default risk as become a challenge since the inception of the financial industry. Datamining is a capable aspect of analysing data aimed to remove beneficial information from incredible numbers of multifaceted dataset. In this study an effective prediction technique that helps the financial industry to predict the credit eligibility for customers applying for loan is proposed using the Chi-Square feature selection model and (SVM-RBF) model. Preceding to constructing the model, the data set is pre-processed, reduced and deliver effective predictions. The results shows that chi-square and SVM-RBF model accuracy was 80.29% and compared with the-state-of-art.

Churn Prediction, Bank Credit, SVM-RBF, Chi-square, Datamining, Credit scoring

The paper assess the impact of total lockdown in Osun Sate, South-West region of Nigeria to reduce the rate of spread of COVID-19 on the Socio-Economic lifestyle of its residents. Date were collected during the luck down by sending out randomly throughout the 3 senatorial district of the State structured questionnaire using goggle form. Information obtained was analysis using SPSS. The study revealed that the socio-economic life style of Osun resident was significantly impaired with secondary impact of possibly reduction in National GDP of the nation.

Economic impact, Corona virus, residents

The corona virus pandemic is one of the largest global health crises in recent times. Several studies are being proposed to demonstrate how Machine Learning techniques can be used for the prediction of the disease. While some of the studies focus on the prediction of the diseases in patients, some other ones focus on predicting its mode of spread across countries of the world. Corona virus was formally announced in China in December, 2019. Since then, the disease has been spreading in large number. More than 3.4 million people have been infected as at 3rd May, 2020. As parts of the efforts to help in predicting the presence of the disease in patients as well as its geographical spread, some Machine Learning-based techniques are being proposed. This study first explores the various promises of Machine Learning as a sub field of Artificial Intelligence in the prediction of corona virus disease. We provided information on how Machine Learning approaches can aid for predictive analytics. We equally emphasized the use of comprehensive and reliable datasets for reliable predictive modeling of the virus. However, since the disease is novel, it is of interest to study some of the datasets that are evolving. This is because having representative datasets for training and testing predictive COVID-19 models is very important. It is believed that future researchers in the area of corona virus classification and prediction can benefit immensely from the findings of this study.

COVID-19, COVID-19 prediction, Machine Learning-based Predictive Models

Crude oil markets have been subjected to shocks and consequently have been highly volatile. Demand and supply shocks cause large movements in oil prices, which are followed by a dynamic response in both energy demand and supply and in the energy exploration and development activities. Modeling crude oil markets is of paramount importance, not only because of the influence of energy on macroeconomic activity but also because of the role of energy in the investment plans of households and firms. Energy cost and efficiency have become a prime concern in these plans. The study attempted to model Nigeria crude oil on gross domestic product (GDP) using Structural Equation Modelling (SEM) for period 1979 to 2013. The time plot revealed that there was increase of crude oil price overtime. In the same vein, the plot for oil production indicates fluctuation and therefore instability. The time plot for oil export showed that there was sharp decline of export overtime. The plot for real gdp indicated that increase in the real gdp overtime, but an abrupt decline from 2010. To avoid spurious regression, the data were tested for stationarity using the Augmented Dickey-Fuller test and the test indicated that the variables were all stationary. This finding establishes that about 53.8% of the total variation in oil export is jointly explained by oil production, crude oil price and real gdp. In the model, oil production, price of crude oil and inflation rate were said to insignificant, while real gdp is statistically significant. Collectively, the all the variables are said to statistically significant. The Durbin-Watson statistic indicates that there is no serial autocorrelation among the variables. Oil export would continue to be under upward pressure so long inflation rate, real gross domestic product and oil production keep falling. The model was examined for long run relationship using cointegration techniques and the study indicates that there is long run disequilibrium (no long relationship). In conclusion, Oil export operates in the demand function and may arise from shocks to real GDP, and inflation rate. Oil production shocks operate in the supply function and may arise from shocks to actual or capacity output.

Structural Equation Modelling (SEM), modelling, Augmented Dickey-Fuller test

An autoregressive (AR) model predicts future behavior based on past behavior. It is used for forecasting when there is some correlation between values in a time series the values that precede and succeed them. Goal programming (GP) model is a simple extension and modification of the linear programming technique that provides a simultaneous solution of a system of complex objectives rather than a single one. The study proposes goal programming as a method to estimate the parameters of the autoregressive model. The findings indicates that in both data sets, the Mean Absolute Percentage Error (MAPE) values using the GP prediction were lower than those obtained using the AR(1) model. Based on these results, it can be deduced that prediction equations obtained using the GP approach were more accurate than those obtained using the AR(1) model. This is because, using the GP approach, the problem can be restated as to minimize the sum of absolute errors rather the sum of squares of the error in the case of the AR(1) model.

Method of least squares, outliers, goal programming