This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: IEOR 4404 Assignment #10 Solutions Simulation December 16, 2008 Prof. Mariana Olvera-Cravioto Page 1 of 3 Assignment #10 Solutions 1. Based on MATLAB output, we choose the following two candidate distributions: • Normal • Lognormal We first use the Kolmogorov-Smirnov test with the Normal CDF, with μ and σ as given by the “Edit Fit” window. The corresponding output is : H = 0 ,p = 0 . 1066, which means that we cannot reject the null hypothesis, at the 95% confidence level. . Then, we use the same test with the Lognormal CDF, with μ and σ as given by the “Edit Fit” window. The corresponding output is : H = 1 and p = 5 . 0704 × 10- 20 , which means that we should reject the null hypothesis. Therefore, the best distribution is the normal distribution. The MLE estimators for its parameters for μ and σ 2 are 1.30438 and 0.409052, respectively. 2. We consider the following two distributions: • Poisson • Binomial with N = 4 (which is the maximum data point observed) We can call the Chi square hypothesis test in MATLAB using the following commands: [h,p] = chi2gof(Data2,‘cdf’,@poisscdf(p,1.492)) for the Poisson, and [h,p] = chi2gof(Data2,‘cdf’,@binocdf(b,4,0.373))[h,p] = chi2gof(Data2,‘cdf’,@binocdf(b,4,0....
View Full Document
- Spring '08
- Normal Distribution, Null hypothesis, Binomial distribution, Likelihood function, mle