Hidden Markov Model

E880217

Markov model generative model probabilistic graphical model statistical model time series model

A Hidden Markov Model is a statistical model that represents systems with unobserved (hidden) states generating observable outputs, widely used for sequence analysis tasks such as speech recognition, bioinformatics, and natural language processing.

Try in SPARQL Jump to: Surface forms Statements Referenced by

Observed surface forms (1)

Surface form	Occurrences
hidden Markov models	1

Statements (63)

Predicate	Object
instanceOf	Markov model ⓘ generative model ⓘ probabilistic graphical model ⓘ statistical model ⓘ time series model ⓘ
appliedIn	computational biology ⓘ finance ⓘ natural language processing ⓘ robotics ⓘ signal processing ⓘ
assumesProperty	conditional independence of observations given states ⓘ first-order Markov property on hidden states ⓘ
hasAlgorithm	Backward algorithm NERFINISHED ⓘ Baum-Welch algorithm NERFINISHED ⓘ Expectation-Maximization for parameter learning ⓘ Forward algorithm ⓘ Forward-Backward algorithm NERFINISHED ⓘ Viterbi algorithm NERFINISHED ⓘ
hasComponent	emission probability distribution ⓘ initial state distribution ⓘ set of hidden states ⓘ set of observable symbols ⓘ state transition probability matrix ⓘ
hasHiddenStates	yes ⓘ
hasObservableOutputs	yes ⓘ
hasVariant	Gaussian mixture Hidden Markov Model NERFINISHED ⓘ continuous Hidden Markov Model NERFINISHED ⓘ discrete Hidden Markov Model NERFINISHED ⓘ hidden semi-Markov model ⓘ higher-order Hidden Markov Model NERFINISHED ⓘ input-output Hidden Markov Model NERFINISHED ⓘ
introducedIn	1960s ⓘ
learningType	supervised learning (when state sequences are known) ⓘ unsupervised learning (for parameter estimation) ⓘ
models	sequences of observations ⓘ temporal processes ⓘ
notableContributor	Leonard E. Baum NERFINISHED ⓘ
parameter	emission probabilities ⓘ initial state probabilities ⓘ transition probabilities ⓘ
relatedTo	Kalman filter NERFINISHED ⓘ Markov chain NERFINISHED ⓘ conditional random field NERFINISHED ⓘ dynamic Bayesian network NERFINISHED ⓘ
supportsTask	decoding most probable state sequence ⓘ likelihood computation ⓘ parameter estimation from data ⓘ
typicalAssumption	finite number of hidden states ⓘ stationary transition probabilities ⓘ
usedFor	activity recognition ⓘ anomaly detection in sequences ⓘ bioinformatics sequence analysis ⓘ gene prediction ⓘ handwriting recognition ⓘ language modeling ⓘ machine translation (classical approaches) ⓘ named entity recognition ⓘ part-of-speech tagging ⓘ protein secondary structure prediction ⓘ sequence labeling ⓘ speaker diarization ⓘ speech recognition ⓘ time series segmentation ⓘ

Referenced by (2)

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

Viterbi algorithm → operatesOn → Hidden Markov Model ⓘ

Gibbs sampling → usedIn → Hidden Markov Model ⓘ

this entity surface form: hidden Markov models