O'Reilly logo

Practical Predictive Analytics by Ralph Winters

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

Transforming service calls to a binary variable

Variables with a higher number of levels will often be difficult to manage even though there may be statistically significant differences shown in the curves. This can be due to the smaller sizes of the groups. Rather than try to analyse all the groups at once, it often makes more sense to first find a cutoff point that collapses the variable into a binary outcome.

For example, running the survdiff() function on the number of service calls (ranging from 0 to 5) shows a significant difference between the individual survival curves:

> survdiff(SurvObj ~ Xservice.calls, data = ChurnStudy)Call:survdiff(formula = SurvObj ~ Xservice.calls, data = ChurnStudy)                    N Observed Expected (O-E)^2/E (OE)^2/V ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required