Ideas around modeling risky to be bankrupt companies

msacs09

Hi Experts,

Im trying to come up with the list of features that would deem a company not worthy to invest. What would be the features that would deem a company too risky (i.e) in their path to bankruptcy.  

I have have 2 categories of companies one is late stage, which has decent financial info and the othet is early stage, where we have a challenge that not too much financial info, since there are early stage companies who has little or no revenue. As of now my key data set for early stage is whatever minimal public data avilable form curnchbase etc.,  

my my goal is to come up with a risky feathers for a given company and rank them in the worst order high risky, medium risky and low risky. 

Thank you in advance for your valuable time and info 

Telcontar120

I would work on building two different models to start, one for the early stage and one for the late stage.  The available predictors and the relationships are both likely to be different.
You might want to start with something simple like Naive Bayes and Decision Tree, which are pretty interpretable models before you move onto more complex ones like SVM or neural nets.