Suppose that we have a set of observations, 1 to n , each with measurements on $p = 1 $ feature, $X$ ( i . e . , we have one predictor and one target variable ) . We assume that X is uniformly ( evenly ) distributed on [ 0 , 1 ] . Associated with each observation is a response value Y . Suppose that we wish to predict a test observation's response using only observations that are within 5 % of the range of X closest to that test observation. For instance, in order to predict the response for a test observation with X = 0 . 6 , we will use observations in the range [ 0 . 5 7 5 , 0 . 6 2 5 ] .