Financial services should include risk management in data mining models

Financial services data mining models should incorporate a level of calibration to manage risk enterprise-wide, said Bart Beasens, a professor at Katholieke Universiteit Leuven in Belgium.

Financial services data mining models should incorporate a level of calibration to manage risk enterprise-wide, said Bart Beasens, a professor at Katholieke Universiteit Leuven in Belgium.

Speaking at today's SAS A2010 analytics conference, Beasens said, "More than ever before data mining models steer strategic decisions in financial institutions."

While current models are good at discriminating data - ranking and sorting - analytic models need to provide well-calibrated and accurately projected probabilities based on historical data and future expectations, he said.

Beasens added that IT departments will need to develop and monitor multi-level architectures to support analytic lifecycles, incorporating internal and external data, data mining and, additionally, a dynamic macro-economic model to allow for risk modelling throughout an organisation.

Read more on IT risk management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

-ADS BY GOOGLE

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close