Re: [問題] 如何產生具有負相關性的多維常態

看板Statistics作者 (偶素米蟲)時間16年前 (2010/03/23 23:09), 編輯推噓0(001)
留言1則, 1人參與, 最新討論串2/2 (看更多)
※ 引述《newlittle (小新)》之銘言: : 之前正相關的作法是將 covariance-variance matrix : 利用 cholesky decomposition 分成上下相同的半矩陣 : 之後藉由標準常態值 即可產生具有相關性的多維常態 : 可是現在要產生具有負相關性的值 : 例如其 covariance-variance matrix 為 : 1 -0.5 -0.5 : -0.5 1 -0.5 : -0.5 -0.5 1 : 因為他不是正定矩陣 所以無法使用cholesky decomposition : 這樣一來該如何產生呢? : 謝謝 This is an interesting question. We can generate arbitrary positive equicorrelated exchangeable random variables, but not negative ones. In fact, if I remember correctly, the smallest negative correlation coefficient allowed is -1/(k-1), where k is the number of random variables. One way to generate the most negative equicorrelated exchangeable r.v.'s is to generate k i.i.d. r.v.'s, say X1,..., Xk, then (X1 -m, ..., Xk-m), where m is the average of these k r.v.'s, will yield the smallest correlation coefficient. -- ※ 發信站: 批踢踢實業坊(ptt.cc) ◆ From: 166.137.8.145

03/25 12:43, , 1F
thanks for your help
03/25 12:43, 1F
文章代碼(AID): #1BgDcIbX (Statistics)
文章代碼(AID): #1BgDcIbX (Statistics)