Title :
Constructing Bayesian networks to predict uncollectible telecommunications accounts
Author :
Ezawa, Kazuo J. ; Norton, Steven W.
Author_Institution :
AT&T Bell Labs., Murray Hill, NJ, USA
fDate :
10/1/1996 12:00:00 AM
Abstract :
The complexities of building models that can predict whether a customer account or transaction is collectible are greater than most current learning systems can handle. The authors describe software that builds Bayesian network models for such predictions. They also examine how varying model parameters and hence model structure can affect predictive accuracy
Keywords :
Bayes methods; learning systems; neural nets; prediction theory; Bayesian networks; customer account; model parameters; model structure; predictions; predictive accuracy; uncollectible telecommunications accounts; Bayesian methods; Buildings; Communication industry; Laboratories; Learning systems; Predictive models; Profitability; Risk management; Telecommunications; Transaction databases;
Journal_Title :
IEEE Expert
