Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
1. Discrete random variables. Let X be a discrete random variable with possible values 1, 2 and 3 which probabilities of these values are: P (X = 1) = 0.5, P (X = 2) = 0.2 and P (X = 3) = 0.3). Answer the following questions: • What is the value of µX = E(X)? • Explain in words what µX represents. 2 = V ar(X)? • What is the value of σX 2 represents. • Explain in wods what σX • What is the value of F (2) = P (X ≤ 2)? • What is the value of P (X > −2)? Define a new random variable Y = 2X 2 + 3. Answer the following questions: • What are the possible values of Y and their probabilities? • What is the value of µY = E(Y )? • What is the value of σY2 = V ar(Y )? • What are the values of Cov(X, Y ) and ρXY = corr(X, Y )? • What does ρXY represent? • Plot FY for all the possible values of Y . 2. Continuous random variables. Let Z be a continuous random variable with a standard distribution N (0, 1). Answer the following questions: • What are the possible values of Z? • What is the value of µZ = E(X)? What is i • Explain in words what µZ represents. • What is the value of σZ2 = V ar(X)? • Explain in wods what σZ2 represents. • What is the expression of density function? • What is the expression of the distribution function? 1 • What is the expression of the mean? • What is the expression of the variance? Define a new random variable Y = 3Z + 2. Answer the following questions: • What are the possible values of Y ? • What is the expression of the density function (fY ) of Y ? • What is the expression of the distribution function (FY ) of Y ? • What is the value of µY = E(Y )? • What is the value of σY2 = V ar(Y )? • What are the values of Cov(Z, Y ) and ρZY = corr(Z, Y )? • Plot FY for all the possible values of Y . 3. Random variables and R. Intall R and RStudio in your computer, google it! • Type in the console in the left down screen help.search(”normal distribution”). • Type help (rnorm). • Type ?rnorm • How is the binomial distribution? Simulate a sample fo 200 value of a Bernoulli variable with probability of success 0.3 (W ∼ Bernoulli(p = 0.3)). • What is the value of µW ? How do you estimate it in R? 2 ? How do you estimate it in R? • What is the value of σW Simulate 200 values of Z and Y of exercise 2. • Plot the density and distribution functions of Z and Y. • Estimate their mean, variance and covariance. 4. Conditional probabilities. We have two discrete variables X and Y . X can take values 1, 2, 3 corresponding to three hair styles blonde, dark and bold. The variable Y can take two values 1 and 0 corresponding to female and male. We know that these variables probability table is: 2 X/Y Female (1) Male (0) Total Blonde (1) 0.22 0.10 0.32 Dark (2) 0.29 0.18 0.47 Bold (3) 0.01 0.20 0.21 Total 0.52 0.48 1 • What is the conditional probability P (X = 1|Y = 0)? • What is the P (X = 1)? • P (Y = 0)? • What is the joint probability P (X = 1, Y = 0)? • Are X and Y independent? 5. Linear models, matrix algebra and R. • Create a script called ”myfirstR.R” with the following commands: Y=numeric(2) X=matrix(1, 2,2) X[1,2]=3 X[2,2]=4 beta= c(2, 3) Y=X%*%beta det(X) Y solve(X, Y) solve(X)%*%Y library(fBasics) inv(X)%*%Y • Understand what each command does. • Now run the following commands: 3 X2= matrix(c(1, 3, 2, 6), nrow=2, byrow=T) det(X2) beta2=solve(X2)%*%Y • Why cannot we find a solution to this problem? • Click in the menu ”Session/Set Working directory/To Source File Location”. What happens? • Put your cursor next to the first line of your code and click on the button ”Run” on the right side of this screen. • Click the button ”Source” instead. 6. Create a script (File/New/R script) with the following commands: X=matrix(1, 1000, 3) X[,2]=rchisq(1000, df=2) X[,3]=rnorm(1000, 1, 0.6) epsilon=rnorm(1000, 0, 0.2) beta=c(2, -3, 0.4) Y=X%*%beta+ epsilon solve(X)%*%Y solve(X%*%X)%*%(X%*%Y) solve(X%*%t(X))%*%(t(X)%*%Y) solve(t(X)%*%X)%*%(t(X)%*%Y) model.1<-lm(Y~X) summary(model.1) • Understand what each command does and why some of them fail. • Interpret the results in model.1. 7. Maximum likelihood: Given a sample of n independent identically distributed observations y1 , y2 , . . . , yn , normally distributed yi ∼ N (µ, σ 2 ). • What is the joint density function of this sample? • What is the likelihood function of this sample? 4 • What is the log-likelihood function of this sample? • What is the score vector of the log-likelihood function? • What is the Hessian? 5