Case study in data analytics

Case study in data analytics

1. The advent of data analytics

   • the use of analytical tools that allow the user to extract useful knowledge and quantify the factors that impact events.
   • the explosive growth of data in our times combined with the advanced computational power that renders data analytics a key tool across all businesses and industries
   • Distributed file systems
   • Database management
   • Case study

2. Credit scoring

   • typically used in the decision-making process of accepting or rejecting a loan
   • A credit scoring model is just one of the factors used in evaluating a credit application. Assessment by a credit expert remains the decisive factor in the evaluation of a loan.

3. Data quality

   • This provides a first insight into the patterns inside the data, but also an insight on the trustworthiness of the data itself.
   • In order for the model to be able to make accurate forecasts it needs to see enough examples of what constitutes a default.
   • The data can contain numerical variables or categorical ones
   • it is often the case that outliers are a result of system error.
   • There is a standard framework around QA which aims to provide a full view on the data quality in the following aspects: Inconsistency, Incompleteness, Accuracy, Precision, Missing / Unknown.

4. Model development

   • the cases that the client misses three payments in a row, or, that the sum of missed payments exceeds a certain threshold.
   • The aim of the credit scoring model is to perform a classification: To distinguish the “good” applicants from the “bad” ones.
   • The first phase in the model development requires a critical view and understanding on the variables and a selection of the most significant ones. Failure to do so correctly can hamper the model’s efficiency.

5. Model performance

   • The aim of the modelling exercise is to find the appropriate coefficients βi for all i=0,1,…,n that lead to a model that has two main desirable properties: 1. It fits the existing data 2. It can produce the correct probability of default on data that it has not seen before
     • The first of these requirements is called “Goodness-of-fit” while the second is called “Predictive Power”
   • Goodness of Fit
   • Predictive Power
   • Confusion Matrix

6. Model refinements

   • There is a number of assumptions underlying the logistic regression model
     • The most important of these are:  Linearity in the explanatory variables  Absence of interactions among explanatory variables

7. Model interpretation

   • One of the most appealing features in logistic regression is the transparency in the interpretation of the model
   • Armed with this result we can provide further guidance to the business by giving the impact of each individual explanatory variable to the client’s forecasted probability of default.