There is an interesting course on Credit Risk Modeling in R on DataCamp.
It covers the classic (machine learning) models to determine the risk in giving people a loan. The subject is simple enough to understand so you do not need to spend a lot of time diving into the subject matter. The course is divided in a 4 modules: Introduction, Logistic Regression, Decisions Trees and Evaluating the Credit Risk Model.
You learn some basics on credit & loans (defaults, expected loss,…). Then you explore the data with some core R functions. You see how to deal with “incomplete” data. In the next chapters they focus on regression (what else?) and decision trees concluding with the strategy curve, ROC and AUC.
Each session is a mix of a short video (a few minutes) with hands-on exercises. The exercises are done in R in the browser, so you don’t need any installation on your local machine. Just like in some other courses, part of the code is already given and you just need to complete. It takes about 4 hours to complete, depending on the time you take for the exercises and playing around with the data and plots.
You can take the first chapter for free, for the other chapters you need to pay the monthly or yearly subscription.
- Although it is a good idea to have an R interpreter in the browser, sometimes you are missing the power and ease of use of for example R-Studio. I submitted too many times code that was not ready yet, because I was “debugging”.
- The data sets can be downloaded but some exercises use data that is not made available.
- There is no way to download video & all course material. That would be great when you are not online.
- Cheap. For this quality this is a real bargain. You pay 25€ per month for full access, 250€ for a whole year!
- You don’t need any installation on your local machine, a browser does the job.
- The quality of the videos: they were short, well-presented, to the point, hi-res and easy to follow.
- Overall fun factor was high: you don’t get bored, the pace of the content is ok, the exercises to the point and not too difficult.
You don’t need any R knowledge, so this an entry-level course. Some background in machine learning might help but the content and exercises do not rely on it.
If you have some spare time and are interested in modeling credit risks in R, then this is for you.