bayesian_logistic: Bayesian estimator for logistic regression with binary...

In philohistoria/CCER: Correct Machine Learning Classification Error in Next-Step Regression

Description

This functions takes binary classification as outcome variables, apply Bayesianestimator, and returned corrected coefficient estimates.

Usage

bayesian_logistic(
  formula,
  data,
  fn = fn,
  fp = fp,
  tp = tp,
  tn = tn,
  fpr = fpr,
  fnr = fnr,
  tau_prior = 0.001,
  working_dir = "",
  n.iter = n.iter
)

Arguments

formula	R formula, such as y ~ x
data	a data.frame object
fn	number of false negatives in validation data
fp	number of false positives in validation data
tp	number of true positives in validation data
tn	number of true negatives in validation data
fpr	false positive rate
fnr	false negative rate
tau_prior	prior values of hyperparameters of Bayesian regression variance
working_dir	directory to store JAGS temporary files
n.iter	number of Bayesian draws

Value

a dataframe contining estimated values, standard errors, z values and p values

philohistoria/CCER documentation built on Dec. 22, 2021, 7:50 a.m.