T*****w 发帖数: 802 | 1 IF we modify the classical problem: symmetric random walk starting from 0,
what's expected time and probability to stop at -a, [-a, b](a>0, b>0).
How to approach it? (A: b/(a+b)
1) shift = +2, -1 (with equal probability 1/2)
2) shift = +1, -1 (with p >1/2)
Now, it is not a martingale and I think change of measure is way to do it.
I read some solutions, but still not very clear..
Thanks.. | b***k 发帖数: 2673 | 2 找到了这些
http://mitbbs.com/article_t/Quant/31249959.html
http://mitbbs.com/article_t/Quant/31259671.html
http://mitbbs.com/article_t/Quant/31258403.html
版上应该还有一些此类问题的讨论,
上次daj还是谁还给总结了一下,其中有个链接到某个朋友的数学blog,解决的非常彻
底。
【在 T*****w 的大作中提到】 : IF we modify the classical problem: symmetric random walk starting from 0, : what's expected time and probability to stop at -a, [-a, b](a>0, b>0). : How to approach it? (A: b/(a+b) : 1) shift = +2, -1 (with equal probability 1/2) : 2) shift = +1, -1 (with p >1/2) : Now, it is not a martingale and I think change of measure is way to do it. : I read some solutions, but still not very clear.. : Thanks..
| T*****w 发帖数: 802 | | p*****k 发帖数: 318 | | T*****w 发帖数: 802 | |
|