请教非线性函数的极大似然估计

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

现有非线性函数的参数要估计，需用极大似然法，使用什么软件可以实现啊

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：极大似然估计 似然估计 极大似然 非线性 极大似然法 函数 非线性 极大似然估计

R can do it.

Type ?nls in R for more details!

Examples from R help

     DNase1 <- subset(DNase, Run == 1)

     ## using a selfStart model
     fm1DNase1 <- nls(density ~ SSlogis(log(conc), Asym, xmid, scal), DNase1)
     summary(fm1DNase1)

     ## using conditional linearity
     fm2DNase1 <- nls(density ~ 1/(1 + exp((xmid - log(conc))/scal)),
                      data = DNase1,
                      start = list(xmid = 0, scal = 1),
                      alg = "plinear", trace = TRUE)
     summary(fm2DNase1)

     ## without conditional linearity
     fm3DNase1 <- nls(density ~ Asym/(1 + exp((xmid - log(conc))/scal)),
                      data = DNase1,
                      start = list(Asym = 3, xmid = 0, scal = 1),
                      trace = TRUE)
     summary(fm3DNase1)

     ## using Port's nl2sol algorithm
     fm4DNase1 <- nls(density ~ Asym/(1 + exp((xmid - log(conc))/scal)),
                      data = DNase1,
                      start = list(Asym = 3, xmid = 0, scal = 1),
                      trace = TRUE, algorithm = "port")
     summary(fm4DNase1)

     ## weighted nonlinear regression
     Treated <- Puromycin[Puromycin$state == "treated", ]
     weighted.MM <- function(resp, conc, Vm, K)
     {
         ## Purpose: exactly as white book p. 451 -- RHS for nls()
         ##  Weighted version of Michaelis-Menten model
         ## ------------------------------------------------------------
         ## Arguments: 'y', 'x' and the two parameters (see book)
         ## ------------------------------------------------------------
         ## Author: Martin Maechler, Date: 23 Mar 2001, 18:48

         pred <- (Vm * conc)/(K + conc)
         (resp - pred) / sqrt(pred)
     }

     Pur.wt <- nls( ~ weighted.MM(rate, conc, Vm, K), data = Treated,
                   start = list(Vm = 200, K = 0.1),
                   trace = TRUE)
     summary(Pur.wt)

     ## The two examples below show that you can fit a model to
     ## artificial data with noise but not to artificial data
     ## without noise.
     x <- 1:10
     y <- x                                  # perfect fit
     yeps <- y + rnorm(length(y), sd = 0.01) # added noise
     nls(yeps ~ a + b*x, start = list(a = 0.12345, b = 0.54321),
          trace = TRUE)
     ## Not run:
     nls(y ~ a + b*x, start = list(a = 0.12345, b = 0.54321),
          trace = TRUE)
     ## End(Not run)      

stata 就可以做 

我觉得matlab也可以做

我觉得matlab也可以做

我觉得matlab也可以做

我觉得matlab也可以做

用R中的nlm函数就可以解决，你可以看看这个函数的具体帮组信息！

acslx也可以做