Number of bins for plotting Calibration and bin classifier scores In the Assignment for plotting calibrated and uncalibrated datasets, you mentioned:
"Display the uncalibrated versus the calibrated model predictions with on the x-axis binned classifier scores (10 bins) and
on the y-axis the empirical class membership probability (i.e., proportion of 1s)"
My question is:
Do we have to use 10 bins for this assignment, even if the optimal number of bins found in the related earlier assignment is not 10?
Also, are the binned classifier scores the mid-points (or however else one chooses to implement that) in a bin ?
Answers and follow-up questions Answer or follow-up question 1
Yes and yes.
Michel BallingsSign in to be able to add an answer or mark this question as resolved.