w*******9
发帖数: 1433 | 1 我用这个去predict survival curve做了很多次了，就是这么做的。类似你variable
selection的方法很多人用，大概现实中效果应该不错吧。 |
|
k*******a
发帖数: 772 | 2 Modelling Survival Data in Medical Research, Second Edition |
|
|
|
o*s
发帖数: 623 | 5 不知道有没有哪位朋友有这本电子书可以分享的，谢谢！
Applied Survival Analysis: Regression Modeling of Time to Event Data, 2nd
Edition
David W. Hosmer, Jr., Stanley Lemeshow, Susanne May
ISBN: 978-0-471-75499-2
416 pages
March 2008, ©2008 |
|
c****6
发帖数: 758 | 6 survival time 用月和年不同的单位对model coefficient estimates 有多大影响，对
最后的predictive scores有多大影响？ |
|
A*******s
发帖数: 3942 | 7 theoretically no...
but if survival time is rounded month/year
two methods would have different numbers of tied observations,
which could impact model estimates |
|
i***l
发帖数: 1656 | 8 正处理数据，来这里问下
Survival analysis 能把几次不同实验的结果 Pool到一起么？
如果能，理论依据 是什么？
多谢 |
|
y**3
发帖数: 267 | 9 For model methods used for repeated events such as counting process,
marginal model,is it theroretically correct to look into future? If so, can
the current available software such as sas R output individual risk
estimates at a specififc time especially at a time beyond the observed time
for each individual pateint level data? Or output the survival time for
individual patient for the event?
Any expert can provide expereince?
Thanks |
|
A*******s
发帖数: 3942 | 10 depends...
for AG models with time dependent variable, you have to also forecast the
future paths of these independent variables.
For frailty models, could plug in the empirical Bayes estimate of the
frailty term into the prediction.
And, more meaningful prediction is on the density rather than survival time.
can
time |
|
y**3
发帖数: 267 | 11 Thanks for the answer. It is very helpful. Will dig mroe aobut those two
models.
Can you elaborate a little for me?
So the models can be used to generate future predictions such as hazard
probabilities for the next month from date of censoring for scoring? Then
what assumption is used? Have you ever to put repeated event survival model
into production tp monitor fututre event?
Btw,I read somewhere the counting process is too easy to look into future.IS
that correct:) |
|
C******n
发帖数: 284 | 12 As far as I understand, you have to estimate a parametric model to predict
future survival time. |
|
y**3
发帖数: 267 | 13 For time-to-event survival model,we may do time -sliced classification
models to predict event(yes or no). For the whole data set, the claims9 in
my data) are censored at different times. I can build different models using
data censored at 1 year, 2 year, 3 years, 4 years etc. For example, for the
2 year model, I use all the claims IDs of age >=2 years and censor them at
2 years. Then use the two years data to find the target and predictors. I
think that is statistically how we do it. Then this... 阅读全帖 |
|
y**3
发帖数: 267 | 14 this is time -to-event data. But we want to perform time-sliced model
instead of survival model. for instance, to build a model at 1 year to
predict the event or not. The set up of the data should censor all the data
beyond 1 year and then to identify event and predictors within 1-year and
then train the model. Not using data beyond 11 year to identify event or not
for the 1-year model?
Anybody confirm? thanks |
|
j*****e
发帖数: 182 | 15 What your doing could be very dangerous.
What does "censer all the data beyond 1 year" mean? You could have changed
the underlying population of interest without recognizing it.
In my experience, the hardest part of survival analysis is to recognize your
actual study population and the sampling scheme. |
|
a********g
发帖数: 651 | 16 现在有data，里面有2个group 1和2，我想plot survival % difference。
就是说 survival1 减去survival2，然后difference vs follow up time 做图，要有
95% CI。找了半天R里面也没找到合适的package和function。
请大牛指点哈。
谢谢 |
|
|
a********g
发帖数: 651 | 18 survdiff will not work. you can get p value from it, but not the survival
rate difference.
Thanks though. |
|
a********g
发帖数: 651 | 19 现在有data，里面有2个group 1和2，我想plot survival % difference。
就是说 survival1 减去survival2，然后difference vs follow up time 做图，要有
95% CI。找了半天R里面也没找到合适的package和function。
请大牛指点哈。
谢谢 |
|
|
a********g
发帖数: 651 | 21 survdiff will not work. you can get p value from it, but not the survival
rate difference.
Thanks though. |
|
w*******9
发帖数: 1433 | 22 自己手动吧，survfit算出每个curve的值和variance,你再手动取差值，variance就是
俩variance的和，不过要注意前面的variance得是estimated survival function的
variance,因为默认不是这样的 |
|
a********g
发帖数: 651 | 23 I am using survexp to calculate expect survival curve in R right now, but
found out that the real people life table in R stopped at yr 1990. after yr
1990, there is no real data for life table.
Have anybody had same issue before? how could I solve it?
Thanks a lot. |
|
c**********8
发帖数: 38 | 24 求统计 survival analysis， longitudinal analysis， non parametric， missing
data 审稿 |
|
P*****6
发帖数: 273 | 25 各位大拿：
有人知道为什么验证survival predictive model 用c- index 比较多。有什么更好的
方法吗？或者验证出发点不同的也可以。感觉光c-index不能满足需要。我很想知道单
个sample的预测，和实际的真实差距。当然因为有censored data, 直接比时间可能不
行。
多谢！ |
|
g*2
发帖数: 658 | 26 刚刚从最后一个晚上的night float回到家，虽然明天就要开始第一个floor, 心情还是
很轻松，发文庆祝！！
从6月22号晚开始到今天，只休息了5个周日晚上。这40天真的是我这辈子最最艰难的时
刻，从来没有这么stressful， 从来没有这么不愿意去上班， 多少次想把beeper扔了
，说我不干了，多少次想说我要quit。现在终于结束了。
senior resident 和 attending doctor 的 一句话：good job, 让我觉得survive了
intern 第一个月night float感觉真好。
这个月犯了无数的错误，也学了很多很多。 intern 在我们医院就是doctor +
phlebotomist (draw a lot blood) + escorter (transport pts) + nurse (do iv
push, monitor vs). 现在接电话几乎还会习惯的说this is dr. gg2, anyone paged
me? 买了benadryl,今天晚上吃，把生物钟调过来。 |
|
g*n
发帖数: 538 | 27 各位同学，同志，朋友，乡亲们，如今世界动荡不安。
中东烽烟四起，金三胖踪迹成疑，香港占中如火如荼，南海东海剑拔弩张，依博拉到处
传播，NFL丑闻不断。
作为老中医生，我们可以为世界做些什么？应该是什么也做不了。
可是我们都经过了,或正在经历苦难的residency 还有fellowship。请把你的经验写出
来，给后来人一个提示，让更多的老中顺利完成training！
经疯医提议，全体版务支持，现在医疗实践版举办征文活动。主题是 Residency，
fellowship survival 101。 请有志，有识，有闲，有经验的四有朋友们发言。
内容自选，标题不限。想上十大可以耸人听闻，不想的标题前加bssd.
包子已经准备好了，发文就有大包子，火帖有超级大包子。各科的同学们展开行动吧。
活动时间：即日起一个月，看情况增减
活动奖励办法：由于各位医生时间有限，同时本次活动也是为后来人提供帮助，属于公
益性质，为鼓励积极性，准备对发文的重奖。 回帖20以内的文章奖100伪币，20-40 奖
200 伪币，40-60 奖300伪币， 60以上奖400伪币。 |
|
J*********4
发帖数: 1274 | 28 Very nice! Even though not the same specialty, many survival strategies are
universal rule to go by anywhere anyplace.
非常感谢！ |
|
p********r
发帖数: 3799 | 29 okay, 四月分开始了，算是training season开始了，racing season 就要到了。我就
来说说我3个月的plan.
现在的状态是
1）run
现在是每个礼拜跑3次，50～55mpw。基本上是2*2hr+3hr，14~15miles/2hr and
22miles/
3hr. 不能说是全easy pace, 但是感觉不吃力。下一步就是加一些tempo之类的速度练
习。还有试了一次跑步上班，感觉不错，以后可以用这个來作double的练习。现在是2*
7mile，打算慢慢加到2*10。
2） bike
现在还是trainer 阶段。4hr/week, 基本上是2*2hr的样子。估计很快就可以骑室外了
，不过从时间上考虑，可能只能周末的long ride练练骑车了。
平时骑车上班，没有什么量。
3）swim
2*1.5hr/week, 都是长距离,一次2500y～2800y左右。现在基本上没怎么改进技术上的
问题，都是游长距离。希望可以每个礼拜加一次游泳，半个小时，专门那来练drills.
4）strength training
现在完全凭心情，偶尔练练core, 举哑... 阅读全帖 |
|
d******0
发帖数: 22800 | 30 铁哥啊，把你的survival plan也爆下光吧。。男版的打铁plan，我琢磨着估计更猛！
我就是想见识一下，赛前大家都是如何魔鬼训练的。 |
|
I***i
发帖数: 14557 | 31 http://www.newscientist.com/article/dn23288-birds-evolve-shorte
Birds evolve shorter wings to survive on roads
Nature, red in wheel and fender. Birds in Nebraska have evolved shorter
wings, which may help them avoid dying on roads by taking off quickly and
darting away from cars.
Eighty million US birds are killed by traffic each year. Cliff swallows (
Petrochelidon pyrrhonota) have taken to nesting on road bridges, so may be
especially vulnerable.
Charles Brown of the University of Tulsa in Okl... 阅读全帖 |
|
c***z
发帖数: 6348 | 32 Great use.
For example, user attrition using survival analysis and sentiment analysis
using naive Bayes. |
|
P*****6
发帖数: 273 | 33 谢谢，
那我要把survival analysis深入研究一下。
naïve bayes 好像是 classification 的方法。很简单易用，有效。但其难度和
理论应用的广度比真正的bayesian analysis差远了。通常不归在bayesian analysis。
当初学bayesian analysis， 要先学EM 和 sampling, 洋洋上百页的讲义，连
empirical bayes 都只有一页。naïve bayes 根本没提。 但是bayesian 方法即
使在统计中都是小众。它的广度只是理论上的。
这个版有点冷清，全靠你了。其实，我觉得data science很值得交流，就算是需要不少
编程知识，大家只要看清方向，多交流，也会提高比较快，用不了一两年就可以实用了 |
|
d******4
发帖数: 132 | 34 Could you give more details about how to use survival analysis on user
attrition?
thanks |
|
A*******s
发帖数: 3942 | 35 理论上就是Poisson process，semiparametric inference那些东西
不过应用起来知不知道这些理论关系不大
不大熟悉survival model在reliability方面的应用，不好说。 |
|
P*****6
发帖数: 273 | 36 我猜您是想说paramatric survival analysis, 比如假设weibull distribution？
这个说不定对customer attrition 真的很有用。
X) |
|
P*****6
发帖数: 273 | 37 大牛，受教了， 我倒是不怀疑牛人们会有办法。就是不知道啥办法。
Cox model 我就知道个adjusted survival curve.
it？ |
|
|
w*******y
发帖数: 60932 | 39 Meritline has the 1 PackWilderness Survival Flint Fire Steel Starter Striker
. Last time it was on people were jumping up and down.
Coupon Code :MLCK600702072853NL1 Only 999 coupons left
|
|
|
|
w*******y
发帖数: 60932 | 42 Meritline
Wilderness Survival Flint Fire Steel Starter Striker $1.49 w/ code
MLCK600702081360NL1
other deals
5-Piece Stylus for Nintendo DS $0.29 w/ code MLCK267241081382NL1
Mini Foldable High Quality Universal USB Travel Charger $0.89 w/ code
MLCK259194081372NL1
20 Sheets Glossy Photo Paper $0.49 w/ code MLC175064081367NL1
*deals start at 5 AM PST*
|
|
|
w*******y
发帖数: 60932 | 44 Meritline
8-in-1 Outdoor Mini Survival Kit: Including LED Flashlight, Compass,
Thermometer, Magnifier, Whistle
$2.99 - $2 code MLCK400061091075AL1
= $0.99 with free shipping
|
|
|
|
w*******y
发帖数: 60932 | 47 New 7-in-1 Survival Camping Tool - Functions as a Hatchet, Spade, Saw,
Hammer, Bottle Opener, Wrench, Compass, And Much More $11.98 shipped
Link:
http://home.dailysteals.com/
|
|
|
|
