SAS : Brier Score for Model Calibration


SAS Analytics : Practical SAS, Statistics & Analysis Course


Data Science: Machine Learning A-Z: Hands-On Python & R In Data Science

The Brier score is an important measure of calibration i.e. the mean squared difference between the predicted probability and the actual outcome.
Lower the Brier score is for a set of predictions, the better the predictions are calibrated.
  1. If the predicted probability is 1 and it happens, then the Brier Score is 0, the best score achievable.
  2. If the predicted probability is 1 and it does not happen, then the Brier Score is 1, the worst score achievable.
  3. If the predicted probability is 0.8 and it happens, then the Brier Score is (0.8-1)^2 =0.04.
  4. If the predicted probability is 0.2 and it happens, then the Brier Score is (0.2-1)^2 =0.64.
  5. If the predicted probability is 0.5, then the Brier Score is (0.5-1)^2 =0.25, irregardless of whether it happens.
By specifying fitstat option in proc logistic, SAS returns Brier score and other fit statistics such as AUC, AIC, BIC etc.
proc logistic data=train;
model y(event="1") = entry/ outroc = rocstats;
score data=valid out=valpred fitstat;
run;
Coursera Data Science

SAS Tutorials : 100 Free SAS Tutorials


Statistics Tutorials : 50 Statistics Tutorials

Get Free Email Updates :
*Please confirm your email address by clicking on the link sent to your Email*

Related Posts:

3 Responses to "SAS : Brier Score for Model Calibration"

  1. Hi Deepanshu,

    I tried getting Brier score using fitstat option but seems I didnt get it.. Can you please place a snapshot here? Do i need to add any other option to get this score.

    ReplyDelete
    Replies
    1. Which version of SAS you are using? The fitstat option in score statement is available in SAS versions starting from SAS 9.3

      Delete
  2. hello,
    how is it possible to perform easily a calibration plot (observed probabilities as a function of the expected one) for a logistic model ?
    best regards

    ReplyDelete

Next → ← Prev