This is some work that was done for a class project. A simple credit scorecard model built using hadoop map-reduce.
This was done as a excersize to see if map-reduce could be used to build and train credit scorecard models. A typical credit scorcard model is a logistic regression trained on data where the independent variables are binned by their weights of evidence. (WoE) The dependent variable is usually a binary indicator indicating the presence of a credit default. The accompanying report details the entire process.