A. Which of the following models are linear? If a model is non-linear, indicate the non-linear term.
I. y = ̠₀(̠₁x₁ + ̠₂x₂) + ̠
II. y = ̠₁(x₁ – x̄₁) + ̠₂(x₂ – x̄₂) + ̠
III. y = exp(̠₀ + ̠₁x) + ̠
IV. y = ̠₀ + ̠₁x₁" + ̠₂x₂" + ̠
B. Write the corresponding mathematical models for the following linear models in "R":
I. lm( y ~ x1 : x2 )
II. lm( y ~ I(1/x) + I(x^3) )
C. Let Y₁ ∼ N(2, 2) , Y₂ ∼ N(−4, 2) , Y₃ ∼ N(−2, 2) and Y₄ ∼ N(2, 4) be independent random variables. Answer the following questions and explain your workings.
I. Find the distributions of X₁ = 2Y₁ + Y₂ and X₂ = Y₁ − 2Y₃
II. Construct two independent random variables Z₁ = Z₁(Y₁, Y₂, Y₃, Y₄) and Z₂ = Z₂(Y₁, Y₂, Y₃, Y₄) such that Z₁ ∼ χ²₃ and Z₂ ∼ χ²₁
III. Construct a random variable W = W(Y₁, Y₂, Y₃, Y₄) such that W ∼ F₃,₁
D. Briefly explain what are the standard assumptions on random errors in the simple linear regression model.