Which of the following is true about local linear regressions? Select all that apply. As the number of observations increases, the interval size can decrease without compromising your estimates. They are always preferable to kernel regressions. The derivative of the function on a given interval is the coefficient of the weighted linear regression on that interval. They are preferable to kernel regression for producting predicted values at the boundaries. unanswered