预测 Mack ChainLadder

问题描述

我正在尝试在另一个三角形上使用 Mack Chainladder 的开发系数，但无法使用 Predict.TriangleModel 使其工作。这是我正在尝试做的一个例子：

示例：

#Triangle 1
RAA <- t(matrix(c(
  5012,8269,10907,11805,13539,16181,18009,18608,18662,18834,106,4285,5396,10666,13782,15599,15496,16169,16704,3410,8992,13873,16141,18735,22214,22863,23466,5655,11555,15766,21266,23425,26083,27067,1092,9565,15836,22169,25955,26180,1513,6445,11702,12935,15852,557,4020,10946,12314,1351,6947,13112,3133,5395,2063,0
),ncol=10))


tri <- as.triangle(RAA) #convert from dataframe to triangle
-- steps to clean upper and lower triangle not shown
tri.cum <- incr2cum(tri) #cumulative triangle
wt <- tri.relevent.wts(tri,12)

MCL <- MackChainLadder(tri.cum,est.sigma="Mack",weights = wt,alpha = 1)
MCL$f                       #development factors
MCL$FullTriangle            #cumulative full triangle

#Triangle 2
RAA2 <- t(matrix(c(
  6012,206,4410,6655,2092,2513,657,2351,4133,3063,ncol=10))

tri2 <- as.triangle(RAA2) #convert from dataframe to triangle
-- steps to clean upper and lower triangle not shown
tri.cum2 <- incr2cum(tri2)


# apply model from prevIoUs triangle to new cumulative triangle
a <- predict(MCL,tri.cum2) # predict the new triangle using MCL model
a$FullTriangle

为什么第二个完整三角形仍然显示为与第一个完全一样？

解决方法

暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！

如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。

小编邮箱:dio#foxmail.com (将#修改为@）