# Question: Suppose we use an AR 2 model to predict the

Suppose we use an AR (2) model to predict the next value of a random process based on observations of the two most recent samples. That is, we form

Ẏ [n + 1] = a1Y [n] + a2Y [n – 1]

(a) Derive an expression for the mean- square estimation error,

E [Ɛ2] = E [(Y n +1] –Ẏ [(n + 1]) 2]

(b) Find the values of the prediction coefficients, a1 and a2, that minimize the mean-square error.

Ẏ [n + 1] = a1Y [n] + a2Y [n – 1]

(a) Derive an expression for the mean- square estimation error,

E [Ɛ2] = E [(Y n +1] –Ẏ [(n + 1]) 2]

(b) Find the values of the prediction coefficients, a1 and a2, that minimize the mean-square error.

## Answer to relevant Questions

Suppose we wish to predict the next value of a random process by forming a linear combination of the most recent samples: Find an expression for the values of the prediction coefficients which minimize the mean- square ...Suppose we form a smoothed periodogram of the PSD, As defined in Equation (10.35), using a rectangular smoothing function, Where fΔ is the width of the rectangle. If we want to form the same estimator using a windowed ...Let Where all of the ωn are non- zero constants, the an are constants, and the θn are IID random variables, each uniformly distributed over [0, 2π]. (a) Determine the autocorrelation function of X (t). (b) Determine the ...If the inputs to two linear filters h1 (t) and h2 (t) and X1 (t) and X2 (t), respectively, show that the cross- correlation between the outputs Y1 (t) and Y2 (t) of the two filters is Determine the noise equivalent bandwidth for a filter with impulse response h (t) = bexp(– at) u (t)Post your question