Publications

Dissemination of project results
Publications

• Leave-one-out crossvalidation favors inaccurate estimators
  by Angelika Geroldinger, Lara Lusa, Mariana Nold and Georg Heinze, submitted.
• Separation in Logistic Regression - Causes, Consequences and Control
  Mohammad Ali Mansournia, Angelika Geroldinger, Sander Greenland and Georg Heinze. American Journal of Epidemiology 2018; 187(4): 864–870.
• Artificially generated near-infrared data for classification purposes
  by Vilma Sem, Ana Kolar and Lara Lusa. Chemometrics and Intelligent Laboratory Systems 2018; 172: 100-108.
• What (not) to expect when classifying rare events
  by Rok Blagus and Jelle J. Goeman. Briefings in Bioinformatics 2018; 19(2): 341–349.
• Firth’s logistic regression with rare events: accurate effect estimates AND predictions?
  by Rainer Puhr, Georg Heinze, Marina Nold, Lara Lusa and Angelika Geroldinger. Statistics in Medicine 2017; 36(14): 2302-2317.
• Gradient boosting for high-dimensional prediction of rare events
  by Rok Blagus and Lara Lusa. Computational Statistics and Data Analysis 2017; 113: 19-37.

Talks

• Prediction and explanation in studies with rare events: problems and solutions
  by Georg Heinze, November 2018, Freiburg, Germany.
• Prediction and explanation in studies with rare events: problems and solutions
  by Georg Heinze, at the BNS-AMED Spring Meeting, June 2018, Rotterdam, Netherlands.
• Ridge regression - a solution to separation in logistic regression?
  by Hana Sinkovec, at the "Young Statisticians"-Meeting of the WBS, October 2017, Vienna, Austria.
• Logistic regression with rare events: problems and solutions
  by Georg Heinze, September 2017, Barcelona, Spain.
• Who is afraid of... Bayesian non-collapsibility
  by Georg Heinze, at the CEN-ISBS 2017, August 28 - September 1, Vienna, Austria
• Rare events bias of logistic regression
  by Rok Blagus, at the CEN-ISBS 2017, August 28 - September 1, Vienna, Austria
• C-statistics: should leave-one-out crossvalidation be banned?
  by Angelika Geroldinger, at the CEN-ISBS 2017, August 28 - September 1, Vienna, Austria
• Leave-one-out crossvalidation favors inaccurate estimators
  by Angelika Geroldinger, at the EMR-IBS, May 2017, Thessaloniki, Greece
• The multiple faces of shrinkage
  by Georg Heinze, March 2017, Oslo, Norway
• New modifications of Firth's penalized logistic regression
  by Angelika Geroldinger, at the Seminar in Graz on November 24th 2016, Austria
• Impact of rare events on predicted probabilities from logistic regression
  by Lara Lusa, at the Applied Statistics conference 2016, September 18 - 21, Ribno (Bled), Slovenia
• Rare events bias of logistic regression
  by Rok Blagus, at the Applied Statistics conference 2016, September 18 - 21, Ribno (Bled), Slovenia
• Who is afrain of.... Bayesian non-collapsibility?
  by Georg Heinze, at the Annual Conference of the International Soceity for Clinical Biostatistics 2016, August 21 - 25, Birmingham, UK
• Firth's penalized likelihood logistic regression: accurate effect estimates AND predictions?
  by Angelika Geroldinger, at the International Biometric Conference, July 2016, Victoria Canada
• Firth's penalized likelihood logistic regression: accurate effect estimates AND predictions?
  by Angelika Geroldinger, at the Fourth Joint Statistical Meeting of the Deutsche Arbeitsgemeinschaft Statistik 2016, March 14 - 18, Göttingen, Germany
• Penalized likelihood logistic regression with rare events
  by Georg Heinze, Invited Seminar in November 2015, ISPED, University of Bordeaux, France
• Penalized logistic regression with rare events: preliminary results
  by Lara Lusa, at the Applied Statistics conference 2015, September 20 - 23, Ribno (Bled), Slovenia
• Accurate Prediction of Rare Events with Firth’s Penalized Likelihood Approach
  by Angelika Geroldinger, at the Applied Statistics conference 2015, September 20 - 23, Ribno (Bled), Slovenia
• Challenges in accurate prediction of rare events with penalized likelihood methods
  by Georg Heinze, at the Applied Statistics conference 2015, September 20 - 23, Ribno (Bled), Slovenia

Selected previous publications of members of our groups

• Blagus R and Lusa L. Joint use of over- and under-sampling techniques and cross-validation for the development and assessment of prediction models. BMC bioinformatics 2015; 16:363.
• Blagus R and Lusa L. Boosting for high-dimensional two-class prediction. BMC Bioinformatics 2015; 16:300.
• Blagus R and Lusa L. Improved shrunken centroid classifiers for high‐dimensional class‐imbalanced data. BMC Bioinformatics 2013; 14:64.
• Blagus R and Lusa L. Class prediction for high‐dimensional class‐imbalanced data. BMC Bioinformatics 2010; 11:523.
• Heinze G and Puhr R. Bias‐reduced and separation‐proof conditional logistic regression with small or sparse data sets. Statistics in Medicine 2010; 29(7‐8): 770‐777.
• Heinze G and Dunkler D. Avoiding infinite estimates of time‐dependent effects in small‐sample survival studies. Statistics in Medicine 2008; 27(30): 6455‐6469.
• Heinze G. A comparative investigation of methods for logistic regression with separated or nearly separated data. Statistics in Medicine 2006; 25(24): 4216‐4226.
• Heinze G and Schemper M. A solution to the problem of separation in logistic regression. Statistics in Medicine 2002; 21(16): 2409‐2419.
• Heinze G and Schemper M. A solution to the problem of monotone likelihood in Cox regression. Biometrics 2001; 57(1): 114‐119.

Background literature

• Ambler G, Seaman S and Omar RZ. An evaluation of penalised survival methods for developing prognostic models with rare events. Statistics in Medicine 2012; 31(11‐12): 1150‐1161.
• Bach FR. Bolasso: model consistent lasso estimation through the bootstrap. Proceedings of the 25th International  Conference on Machine Learning, ACM 2008; 307: 33‐40.
• Bull SB, Mak C and Greenwood CMT. A modified score function estimator for multinomial logistic regression in small samples. Computational Statistics & Data Analysis 2002; 39(1): 57‐74.
• Bull SB, Lewinger JP and Lee SSF. Confidence intervals for multinomial logistic regression in sparse data. Statistics in Medicine 2007; 26(4): 903‐918.
• Firth D. Bias reduction of maximum likelihood estimates. Biometrika 1993; 80(1): 27‐38.
• Friedman J, Hastie T and Tibshirani R. Regularization paths for generalized linear models via coordinate descent. Journal of Statistical Software 2010; 33(1): 1‐22.
• Goeman JJ. L1 penalized estimation in the Cox proportional hazards model. Biometrical Journal 2010; 52(1): 70‐84.
• He H and Garcia EA. Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering 2009; 21(9): 1263–1284.
• Henderson R, Jones M and Stare J. Accuracy of point predictors in survival analysis. Statistics in Medicine 2001; 20(20):3083‐3096.
• Hoerl AE and Kennard RW. Ridge regression: Biased estimation for nonorthogonal problems. Technometrics 1970; 12(1):55‐67.
• Jeffreys H. An invariant form for the prior probability in estimation problems. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 1946; 186(1007): 453‐461.
• King G and Zeng L. Logistic regression in rare events data. Political analysis 2001; 9(2): 137‐163.25
• Lee SI, Lee H, Abbeel P and Ng AY. Efficient L1 regularized logistic regression. Proceedings of the 21st National Conference on Artificial Intelligence, AAAI 2006; 1: 401‐408.
• Lin IF, Chang WP and Liao YN. Shrinkage methods enhanced the accuracy of parameter estimation using Cox models with small number of events. Journal of Clinical Epidemiology 2013; 66(7): 743‐751.
• Meinshausen N and Bühlmann P. Stability selection. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2010; 72(4): 417‐473.
• Shen J and Gao S. A solution to separation and multicollinearity in multiple logistic regression. Journal of Data Science 2008; 6(4): 515‐531.
• Simon N, Friedman J, Hastie T and Tibshirani R. Regularization paths for Cox’s proportional hazards model via coordinate descent. Journal of Statistical Software 2011; 39(5): 1‐13.
• Tibshirani R. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society. Series B (Methodological) 1996; 58(1): 267‐288.
• Verweij PJM and Van Houwelingen JC. Cross‐validation in survival analysis. Statistics in Medicine 1993; 12(24): 2305‐2314.
• Zou H and Hastie T. Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society: Series B(Statistical Methodology) 2005; 67(2): 301‐320.
• Zou H. The adaptive lasso and its oracle properties. Journal of the American Statistical Association 2006; 101(476): 1418‐1429