Bayesian logistic regression

E898980

Bayesian model classification method generalized linear model statistical model

Bayesian logistic regression is a probabilistic classification method that models binary outcomes using a logistic link function with prior distributions on the parameters, enabling full Bayesian inference and uncertainty quantification.

Try in SPARQL Jump to: Statements Referenced by

Statements (50)

Predicate	Object
instanceOf	Bayesian model ⓘ classification method ⓘ generalized linear model ⓘ statistical model ⓘ
advantage	automatic tradeoff between fit and complexity ⓘ coherent uncertainty estimates for parameters ⓘ coherent uncertainty estimates for predictions ⓘ incorporation of prior knowledge ⓘ
assumes	conditional independence of labels given features and parameters ⓘ
basedOn	logistic regression ⓘ
belongsTo	Bayesian statistics NERFINISHED ⓘ machine learning ⓘ probabilistic modeling ⓘ
canIncorporate	feature selection via sparsity-inducing priors ⓘ
enables	posterior predictive distributions ⓘ probabilistic classification ⓘ uncertainty quantification ⓘ
extends	frequentist logistic regression ⓘ
handles	regularization via priors ⓘ small sample sizes ⓘ
implementedIn	JAGS NERFINISHED ⓘ PyMC NERFINISHED ⓘ Stan NERFINISHED ⓘ
inferenceMethod	Hamiltonian Monte Carlo NERFINISHED ⓘ Laplace approximation ⓘ Markov chain Monte Carlo NERFINISHED ⓘ expectation propagation ⓘ variational inference ⓘ
likelihoodFunction	Bernoulli likelihood ⓘ binomial likelihood ⓘ
models	binary outcomes ⓘ probabilities of class membership ⓘ
output	posterior distribution over coefficients ⓘ posterior predictive distribution over labels ⓘ
parameterSpace	intercept term ⓘ regression coefficients ⓘ
relatedTo	Bayesian generalized linear models NERFINISHED ⓘ Bayesian probit regression ⓘ
typicalPrior	Cauchy prior on coefficients GENERATED ⓘ Gaussian prior on coefficients GENERATED ⓘ Laplace prior on coefficients GENERATED ⓘ hierarchical priors GENERATED ⓘ
usedFor	binary classification ⓘ credit scoring ⓘ medical diagnosis ⓘ risk prediction ⓘ text classification ⓘ
uses	Bayesian inference ⓘ prior distributions on parameters ⓘ
usesLinkFunction	logistic link ⓘ

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

Gibbs sampling → usedIn → Bayesian logistic regression ⓘ