Assessing loan risks: a data mining case study

The USDA's (US Department of Agriculture) Rural Housing Service administers a loan program that lends or guarantees mortgage loans to people living in rural areas. To administer these nearly 600000 loans, the department maintains extensive information about each one in its data warehouse. As with most lending programs, some USDA loans perform better than others. Basic data mining techniques helped the USDA's Rural Housing Service better understand and classify problem loans.