5. Which of the following estimation techniques doesn't use a prior probability distribution? [5 points] (a) Maximum a posteriori probability (b) Maximum likelihood (c) Conditional expectation (d) Least mean squared error 6. Suppose you have a random variable X with an unknown distribution, it is known E[X] = A, var(X) = o' and you know that X can take positive or negative values. Suppose you need a bound on the chance it takes values that are "far" from its expected value. Which of these tools would be most useful? [5 points] (a) Markov Inequality (b) Chebyshev's Inequality (c) A table of the standard normal CDF (or $ function)