y*****n
发帖数: 5016 | 1 has anyone created any early performance report (very similar to validation-
odds report but shorter performance window) before? how do you select
baselines and measurements? exactly the same as validation-odds, i.e. use
development, OOT, etc as baseline and KS, divergen, roc/gini, pdo, slope,
etc as measurements? or something simpler? | l*****t
发帖数: 8319 | 2 你们公司难道没有模板吗 包括psi。。
validation-
【在 y*****n 的大作中提到】
: has anyone created any early performance report (very similar to validation-
: odds report but shorter performance window) before? how do you select
: baselines and measurements? exactly the same as validation-odds, i.e. use
: development, OOT, etc as baseline and KS, divergen, roc/gini, pdo, slope,
: etc as measurements? or something simpler?
| y*****n
发帖数: 5016 | 3 嘛都没有，必须 build from scratch. 怎么样？这工作刺激吧？哈
【在 l*****t 的大作中提到】
: 你们公司难道没有模板吗 包括psi。。
:
: validation-
| A*******s
发帖数: 3942 | 4 就是copy & paste validation report，
只不过把target variable从performance in long forecasting horizon改成
performance in short forecasting horizon
友情提示，这个report不是重点
重点是用survival model instead of logistic model
要不一般report结果会很难看。
validation-
【在 y*****n 的大作中提到】
: has anyone created any early performance report (very similar to validation-
: odds report but shorter performance window) before? how do you select
: baselines and measurements? exactly the same as validation-odds, i.e. use
: development, OOT, etc as baseline and KS, divergen, roc/gini, pdo, slope,
: etc as measurements? or something simpler?
| c*********y
发帖数: 122 | 5 You probably need to change bad definition if the performance period of your
model is longer.
For example, for acquisition model, if the performance window is 18 months
and bad definition is bk/charge-off/90+, you might want to change to 30+ in
9 months which will be dependent on your portfolio. Other stuffs are the
same.
It seems that there is no risk reporting template in your company. Just
curious how they manage the portfolio. |
|
