Prognostic modelling with logistic regression analysis: a comparison of selection and estimation methods in small data sets.

By Ewout W. Steyerberg, Marinus J. C. Eijkemans, Frank E. Harrell, J. Dik F. Habbema. Statistics in Medicine, 2000

795 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/771956ad7884f8e5178963e3da444837eac9f81d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

An Empirical Comparison of Model Validation Techniques for Defect Prediction Models (2017, methodology): “An influential characteristic in the performance of a model validation technique is the number of Events Per Variable [2], [6], [82], [99], i.”
Sample size for binary logistic prediction models: Beyond events per variable criteria (2018, methodology): “EPV 50 may be needed when stepwise predictor selection with conventional type I error 1⁄4 :05 is applied.(11) Conversely, more recent work suggests that the EPV 10 criterion may be too strict in particular settings, showing several examples where prediction models developed with modern regression sh”
The Impact of Class Rebalancing Techniques on the Performance and Interpretation of Defect Prediction Models (2018, background): “tion and spread of the predicted probabilities [71], [72], [73],”
A strict maximum likelihood explanation of MaxEnt, and some implications for distribution modelling (2013, methodology): “Perhaps the most frequently cited among such qualitative scales is a three-grade scale often referred back to Swets (1988). According to this scale, models are characterised as ʻusefulʼ if 0.”
Predicting mortality among patients hospitalized for heart failure: derivation and validation of a clinical model. (2003, methodology): “Variable selection in multivariable modeling was based on clinical and statistical significance.(29) Weexaminedthestrength and shape of the relationships of continuous variables with the log odds of death using cubic spline plots.”
Regression Modeling Strategies (2003, methodology): “ lasso (shrinkage using L1 norm favoring zero regression coefficients) [186, 179] elastic net (combination of L1 and L2 norms that handles the p > n case better than the lasso) [224] adaptive lasso [221, 205] more flexible lasso to differentially penalize for variable selection and for regres”
Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies: The CHARMS Checklist (2014, background): “However, this approach merely creates two similar but smaller datasets differing only by chance, and generally provides little additional information beyond the apparent performance, and for large datasets, the difference in performance in the development and validation dataset is even negligible, a”