Predictive Modeling Using Transactional Data

| Point of View

Analytics uses past data to forecast or predict future events, providing financial services firms with a strategic capability to be proactive.

Predictive modeling offers the potential for firms to be proactive rather than reactive. Predictive modeling using transactional data poses particular challenges which need to be carefully addressed to create useful models.

In this paper, Capgemini discusses challenges for predictive modeling such as data quality, cohort and trend analysis, model variable definition and model selection.