Performance of Firth-and logF-type penalized methods in risk prediction for small or sparse binary data

Performance of Firth-and logF-type penalized methods in risk prediction for small or sparse binary data