Linked from

The 55 pages that link to Maximum likelihood estimation, each with the reason it gives.

• Molecular phylogeneticsRelated: Likelihood methods compare trees by how well their models explain the observed sequences.

• Bayesian inferenceCompared with: It uses the likelihood alone, unlike Bayesian estimation’s combination of likelihood and prior.

• Newton's methodRelated: Newton-based optimization can compute parameter estimates by solving likelihood equations.

• Bayes' theoremCompared with: Identical to Bayesian estimation when the prior is flat, showing precisely where priors matter.

• Statistical inferenceBroader topic: It is a widely used way to estimate model parameters from data.

• Likelihood functionRelated: Maximizing this function produces the maximum-likelihood estimate.

• Inverse problemRelated: It uses a noise model to define a best-fitting estimate without necessarily imposing a prior.

• Statistical modelRelated: It selects parameter values that make the observed data most likely under the model.

• Ronald FisherRelated: Fisher developed and formalized this method as a foundation for statistical inference.

• Independent and identically distributed random variablesRelated: For i.i.d. data, the likelihood factors into repeated contributions from one distribution.

• Probability mass functionRelated: For discrete observations, the mass function supplies the likelihood being maximized.

• Bayesian statisticsCompared with: It uses the likelihood without combining it with a prior distribution.

• PhylogenomicsRelated: Likelihood methods estimate trees from sequence data under explicit models of evolution.

• Prior probabilityCompared with: Unlike Bayesian estimation, it does not combine a likelihood with a prior distribution.

• Objective functionRelated: Its likelihood, or log-likelihood, is the quantity being optimized.

• StatisticsRelated: It provides a common rule for fitting statistical models to observations.

• EconometricsRelated: Many econometric models are fitted by choosing parameters that make the data most likely.

• Power lawRelated: It provides a principled way to estimate power-law exponents rather than relying on plotted slopes.

• Factor analysisRelated: It is a common method for estimating factor-model parameters.

• Likelihood ratioCompared with: Estimation finds a best-fitting parameter; a likelihood ratio compares hypotheses or models.

• Parameter estimationRelated: Its consistency and finite-sample behavior depend on model assumptions and data conditions.

• Likelihood-ratio testRelated: Each model is fitted by maximizing its likelihood before the comparison.

• Bernoulli distributionRelated: For independent Bernoulli observations, the maximum-likelihood estimate of p is the observed success fraction.

• Frequentist statisticsRelated: Its sampling properties let frequentists assess estimator bias, consistency, and precision.

• Logistic regressionRelated: It commonly fits logistic regression coefficients to observed outcomes.

• Bayesian phylogeneticsCompared with: It typically reports a best-fitting tree rather than a posterior distribution over trees.

• Fisher informationRelated: Its estimator often approaches the inverse-information variance bound in large samples.

• Sufficient statisticRelated: Likelihood factorization can show that maximum-likelihood calculations depend on the sample only through a sufficient statistic.

• Long-branch attractionCompared with: Likelihood methods can outperform parsimony under rate heterogeneity when the model is suitable.

• Method of momentsCompared with: It uses the full likelihood rather than matching selected moments.

• Ordinary least squaresRelated: Under Gaussian errors with constant variance, maximizing likelihood gives the same estimates as OLS.

• Root-finding algorithmRelated: Numerical solvers often find roots of the likelihood's score equations.

• Econometric modelRelated: It estimates models by specifying how the data are distributed.

• Expectation–maximization algorithmNarrower topic: EM developed as a general iterative route to likelihood estimates when direct optimization is difficult.

• Frequentist probabilityRelated: Its repeated-sample properties motivate frequentist evaluations of estimator performance.

• Item response theoryRelated: It is a standard way to estimate item and trait parameters from response data.

• Parameter (statistics)Broader topic: It estimates parameters by making the observed sample most likely under the model.

• Mathematical statisticsBroader topic: It is a central way mathematical statistics turns a specified model into parameter estimates.

• Maximum a posteriori estimationCompared with: Unlike MAP, it does not incorporate a prior distribution.

• Maximum and minimumRelated: It turns parameter estimation into a search for a maximum.

• Poisson regressionRelated: Poisson regression coefficients are commonly estimated by maximizing the count-data likelihood.

• Population parameterRelated: It provides a general recipe for estimating parameters from a specified data model.

• Gutenberg–Richter lawRelated: It offers a principled way to estimate the b-value from earthquake catalogs.

• Numerical optimizationRelated: Numerical optimizers often find parameter estimates when likelihoods lack closed-form solutions.

• Gauss–Markov theoremCompared with: Unlike the theorem’s variance comparison, likelihood methods rely on a specified probability model.

• Trygve HaavelmoRelated: Likelihood-based estimation follows naturally once economic observations are assigned a probability distribution.

• Statistical machine translationRelated: Training commonly estimates translation parameters by maximizing the likelihood of aligned text.

• Lehmann–Scheffé theoremCompared with: A maximum-likelihood estimator need not be unbiased or minimum-variance among unbiased estimators.

• Lotka's lawRelated: It offers a way to estimate the exponent from observed author publication counts.

• Parametric statisticsRelated: It is a standard way to fit parametric models to observations.