The document notes that in Equation 3, the square should be inside the expectation operator, so it should read as the expected value of the error term (y - the predicted value from the model based on the training data) squared, rather than squaring the expected value of the error term.
The document notes that in Equation 3, the square should be inside the expectation operator, so it should read as the expected value of the error term (y - the predicted value from the model based on the training data) squared, rather than squaring the expected value of the error term.
The document notes that in Equation 3, the square should be inside the expectation operator, so it should read as the expected value of the error term (y - the predicted value from the model based on the training data) squared, rather than squaring the expected value of the error term.
The document notes that in Equation 3, the square should be inside the expectation operator, so it should read as the expected value of the error term (y - the predicted value from the model based on the training data) squared, rather than squaring the expected value of the error term.
=========== 10/24: In Eqn 3, the square ("^2") should be inside the expectation, i.e., it should read as: E{error_{1}(S_n)} = E{(y- \hat{f}_{S_{n-1}}(x))^2}