Scoring models are developed by applying statistical procedures to historical data with the objective of predicting a future outcome. They are based on the premise that past behavior is a good predictor of future performance. A scoring model’s purpose is to predict the likelihood that an outcome of interest will occur in a specified time period.
Some major applications include Loan defaults, Pre-screening, Bankruptcy, Limit assignment, Collection modeling and Predicting marketing responses.
Custom scoring models are usually built for a specific client, and tailored to closely match their data and business objectives. Naturally, customization results in a more relevant and accurate prediction tool.