Description

Customer Analytics in Large Organisations

Assignment 2 part 1 – Credit Scoring

Your task is to prepare the data, run an analysis, answer research questions below and write a brief report with your findings.

Dataset

You are provided with personal loans data from Brazil. There are two data sets – Accepted and Rejected. The last one does not have an information about loan outcome as these people were rejected for some subjective reasons. Also, there is a file with data description.

Brazil is a relatively poor country. One of the major challenges in the assignment would be “a low quality” of borrowers in Brazil. You should get models that are sensible and “good enough”.

Research questions

1.       Review data and make decisions what variables to include or exclude from the model. Report and explain your decisions.

2.       Prepare Credit Scoring model. Report decisions made during model development stage, final scorecard, statistical summary of results, discuss quality of the model.
Hint: data set Rejected should have Role “Score”. You can set it in Property menu.

3.       Create a separate “Machine Learning” predictive model – logistic regression, decision tree, neural network, whatever is available in SAS and works best. Alternatively, you can use any other software. Compare its performance to Credit Scoring model. Discuss implications of using each approach.
Hint: in SAS there is no need to import the same data again, you can build a new connection between existing data and some predictive model.

Submission

You must submit a formal report in MS Word or PDF format. Your report will include:

1.       Introduction.

2.       Dataset description.

3.       Credit Scoring model with related discussion. 

4.       An alternative predictive model with related discussion.

5.       Conclusion with comparison of the above two approaches.

6.       Appendix with some extra information, if required.

There is no requirement or limits for word count. Your report should demonstrate completeness in covering all research questions and brevity as no one loves reading long reports. “A picture is worth a thousand words” – use data visualisations to illustrate and support your research findings.

If you have any questions – feel free to ask on the forum. You can discuss this exercise with me and other students. You are encouraged to share ideas but not solutions. Remember about academic integrity.