I am trying to do classification for a Radiomics database on a binary output variable and around 40 variables extracted from MR images.
Right now I am starting pretty easy to find relevant variables in my dataset with the following commands:
Code:
splitsample, generate(sample) split(.8 .2) label define svalues 1 "Training" 2 "Testing" label values sample svalues lasso logit C2 HISTO_Skewness-GLZLM_ZP if sample == 1 lassoknots lassocoef, display(coef, standardized) estimates store lasso lassogof lasso, over(sample)
With the lassocoef command I get some idea about the variables but I have no clue on how to get an idea about accuracy or AUC of the model if applied to my Testing dataset and I cannot find any information in the Stata documentation.
Can anyone help me here on how to get ideas about accuracies, PPW, NPW or even a confusion matrix?
Thanks a lot!
Michael
0 Response to Accuracy / ROC for lasso model?
Post a Comment