So here's the problem:
I am dealing with about 200 time-series of banking accounts that have drawn on a credit at some point. I know the point when they got the credit (let's call this point t(credit) ) and know the account balance from t(0) - t(credit)
Now I want to build a model which tells me, how high the chance for taking a credit is for other accounts that have not taken one so far.
In other words: I want to know at which point a customer takes on a credit.
I know that this is not realy a multiple time series, since the series don't effect each other.
The closest slution so far is described in this article which is not totaly reproducable to me in some parts.
My questions:
What kind of statistical test/ machine learning algorithm can solve the problem?
Is there an R-Package that might help me?
I hope the problem is described clearly enough... sorry for bad english