Browsing by Author "Folorunsho, A. I."
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item EVALUATION OF HOTELLING T2 AND MULTIVARIATE ANALYSIS OF VARIANCE TESTS(Faculty of Sciences, Federal University of Technology Minna, Nigeria, 2016) Oyeyemi, G. M.; Adebayo, P. O.; Folorunsho, A. I.Multivariate analysis of variance (MANOVA) which comprises of Wilks' lambda, Pillai's trace, Lawley-Hotelling trace and Roy's largest root are compared with Hotelling T square when null hypothesis is true. Data were simulated to compare the five (5) test statistics under the two different distributions (Multivariate Gamma and Multivariate Normal), sample size (10, 30, 60, 90, 400, 500, 800 and 1000), number of variables p = 2 and equal and unequal sample size and variance covariance matrix. The comparisons were done at two levels of significant (alpha = 0.01 and 0.05) using power of the test and type I error rate. The results showed that Roy's largest root test statistic is better than all other test statistics considered when sample size are equal but Hotelling T square performed better for unequal sample sizeItem Micronumerosity in Classical Linear Regression(College of Natural and Applied Sciences, University of Port Harcourt, Nigeria, 2015) Oyeyemi, G. M.; Bolakale, A; Folorunsho, A. I.; Garba, M. K.This study studied the problem of micronumerosity in Classical Linear Regression (CLR) in other to prescribe appropriate remedy to the problem if encountered at any CLR analysis. The study is aimed at determining an optimum sample size n*, such that when the number of observations of variables in CLR is greater than (i.e n > n*) then micronemerosity is not a problem. It also suggests means of correcting micronumerosity in CLR. The minimum sample size (n) for a given number of independent variables (p) and level of correlation between the dependent and independent variable(s) were determined. Also, Factor analysis served as the best method of overcoming problem of micronumerosity.Item On Performance of Shrinkage Methods - A Monte Carlo Study(Faculty of Sciences, Federal University of Technology Minna, Nigeria., 2016) Oyeyemi, G. M.; Ogunjobi, E. O.; Folorunsho, A. I.Multicollinearity has been a serious problem in regression analysis, Ordinary Least Squares (OLS) regression may result in high variability in the estimates of the regression coefficients in the presence of multicollinearity. Least Absolute Shrinkage and Selection Operator (LASSO) methods is a well established method that reduces the variability of the estimates by shrinking the coefficients to exactly zero. We present the performance of LASSO-type estimators in the presence of multicollinearity using Monte Carlo approach. The performance of LASSO, Adaptive LASSO, Elastic Net, Fused LASSO and Ridge Regression (RR) in the presence of multicollinearity in simulated data sets are compared using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) criteria. A Monte Carlo experiment of 1000 trials was carried out at different sample sizes n (50, 100 and 150) with different levels of multicollinearity among the exogenous variables (p = 0.3, 0.6, and 0.9). The overall performance of LASSO appears to be the best but Elastic net tends to be more accurate when the sample size is large.