![]() ![]()
The asymptotic significance of each model is less than 0.05, so all are doing
better than guessing.
From the confidence intervals, you can see that model 3 is inferior to the other two
because the entirety of its interval lies below the others.
Models 1 and 2 are nearly indistinguishable, so your model is adopted because
it requires less input to provide a recommendation.
|
|
Area under the Curve |