ocal linear regression is the same as a kernel regression, except that instead of taking a weighted average of observations within an interval as your predicted value, you take the predicted value obtained by running a weighted regression over the set of observations in that interval. In both cases, the weights are given by a kernel