The weights are derived from the fact that the expected number of unique patterns in a bootstrap sample of size N is approximately (1 − e−1)N = 0.632N.
確定！ 回上一頁
