hsstan - Hierarchical Shrinkage Stan Models for Biomarker Selection

Linear and logistic regression models penalized with hierarchical shrinkage priors for selection of biomarkers (or more general variable selection), which can be fitted using Stan (Carpenter et al. (2017) <doi:10.18637/jss.v076.i01>). It implements the horseshoe and regularized horseshoe priors (Piironen and Vehtari (2017) <doi:10.1214/17-EJS1337SI>), as well as the projection predictive selection approach to recover a sparse set of predictive biomarkers (Piironen, Paasiniemi and Vehtari (2020) <doi:10.1214/20-EJS1711>).

Last updated 10 months ago

bayesianfeature-selectionmcmc

3.56 score 6 stars 12 scripts 276 downloads

nestfs - Cross-Validated (Nested) Forward Selection

Implementation of forward selection based on cross-validated linear and logistic regression.

Last updated 2 years ago

feature-selection

3.04 score 2 stars 11 scripts 339 downloads

wevid - Quantifying Performance of a Binary Classifier Through Weight of Evidence

The distributions of the weight of evidence (log Bayes factor) favouring case over noncase status in a test dataset (or test folds generated by cross-validation) can be used to quantify the performance of a diagnostic test (McKeigue (2019), <doi:10.1177/0962280218776989>). The package can be used with any test dataset on which you have observed case-control status and have computed prior and posterior probabilities of case status using a model learned on a training dataset. To quantify how the predictor will behave as a risk stratifier, the quantiles of the distributions of weight of evidence in cases and controls can be calculated and plotted.

Last updated 5 years ago

aucweight-of-evidence

2.00 score 10 scripts 181 downloads