This paper presented Kernel quantile regression in the prediction of inflation rate based on foreign exchange, crude oil price and currency in circulation. Two Radial basis kernel functions namely the Gaussian kernel and the Laplace kernel were applied in the analysis using four different sigma values (5,10,100,1000). Predictions at four quantiles (0.25, 0.50, 0.75 and 0.95) were made and comparison of the predicted values and the observed values were done based on the Root Mean Square Error of Prediction, Relative absolute error, Correlation and Paired T-test. Non-linearity test, normality test, Heteroscedasticity tests were carried out and the results revealed non-linearity, non-normality and presence of heteroscedasticity. The results reveal that the prediction error using the Gaussian kernel function was high for sigma 5,10,100, while that of the Laplace was only high for sigma value 5. The correlation between the observed and predicted was very high for both kernel functions at all quantiles but the paired t-test for the Gaussian kernel showed significant difference all the sigma except 1000, while there was not significant difference using the Laplace kernel at all quantiles. This paper concludes that the Laplace kernel is less sensitive to the sigma value as it makes good predictions at almost the whole considered sigma values but the Gaussian kernel is seen to have good prediction for sigma value 1000 hence it is very sensitive to the sigma value in the kernel function.

Effects of forchheimer on MHD fluid on a rotating cone in the presence of soret and dufor has been conducted in this research. The governing equations are non linear partial differential equations. Using similarity transformations the equations are transformed to ordinary differential equations. The solution obtained on effects of various parameters such as Forchheimer (L), Duffor (Dr). Sorret (Sr), buoyancy parameter (λ) and porosity parameter (έ). It was observed that L, Sr and Dr regulate tangential velocity and appreciate normal velocity. Furthermore, the effects of other parameters over the main physical quantities are provided graphically and also extensively discussed. The problem is solved numerically using inbuilt Runge Kuta shooting technique in Matlab package Bvp4c.

Numerous public sector initiatives to maintain the health of the economy are funded by the federal government. Both one-time purchases and continuous operations require money from the Nigerian government. A capital expenditure is the purchase of new or used fixed assets, whereas an operating expense is the cost of running a firm, which includes the cost of supplies, labor, and depreciation. Data from the areas of administration, sociology, economics, and the community were examined using vector autoregression (VAR). Plots for each of the five sectors were created in order to identify any trends that might exist between them. The trend component was found to be quadratic with respect to both appreciation and depreciation over time. To ascertain how closely related the variables are to one another, the correlation matrix was obtained. It was demonstrated that all the factors have a strong positive association (above 80%). Using ARIMA and VAR models, the performance of the sector was modeled. The best model was chosen for additional investigation after ARIMA and VAR models were created for every one of the economic variables. Using the AIC and BIC selection criterion, the VAR 2 of lag 4 was determined to be the appropriate model for the data set used. Additionally, a projection for the subsequent six years was made using the VAR (2) model. Vector autoregression effectively captures the dynamic nature of interactions (VAR). ADF and KPSS were used to confirm the unit roots. The AIC, BIC, and HQC metrics were used in the calculation of series lag. The VAR(2) model's parameters were estimated.