On the Convergence Properties of the EM Algorithm

Wu, Changbao

doi:10.1214/aos/1176346060

Cited by 2,691 publications

(1,567 citation statements)

References 17 publications

Supporting

Mentioning

1,531

Contrasting

Unclassified

Order By: Relevance

“…, y * (M ) i,mis . Thus, by Theorem 2, the sequence l * obs (θ (t) ) is monotonically increasing and, under the conditions stated in Wu (1983), the convergence ofθ (t) to a stationary point follows for fixed M . Theorem 2 does not hold for the sequence obtained from the Monte Carlo EM method for fixed M , because the imputed values are re-generated for each E-step of the Monte Carlo EM method, and convergence is very hard to check for the Monte Carlo EM (Booth & Hobert, 1999).…”

Section: Therefore (19) Implies (20)mentioning

confidence: 79%

Parametric fractional imputation for missing data analysis

2011

View full text Add to dashboard Cite

SUMMARYParametric fractional imputation is proposed as a general tool for missing data analysis. Using fractional weights, the observed likelihood can be approximated by the weighted mean of the imputed data likelihood. Computational efficiency can be achieved using the idea of importance sampling and calibration weighting. The proposed imputation method provides efficient parameter estimates for the model parameters specified in the imputation model and also provides reasonable estimates for parameters that are not part of the imputation model. Variance estimation is discussed and results from a limited simulation study are presented.

show abstract

Section: Therefore (19) Implies (20)mentioning

confidence: 79%

Parametric fractional imputation for missing data analysis

2011

View full text Add to dashboard Cite

show abstract

“…These two steps will be done iteratively until it converges to a local maximum of the likelihood function (Schafer 1997). A detailed explanation on the convergence properties of EM algorithm can be found in some literature, as an example by Wu (1983).…”

Section: Expectation-maximization Algorithmmentioning

confidence: 99%

Missing Value Estimation Methods for Data in Linear Functional Relationship Model

Ghapor¹,

Zubairi²,

Imon³

2017

JSM

View full text Add to dashboard Cite

Missing value problem is common Data lenyap sering terjadi dalam analisis data kuantitatif. Dengan berkembangnya keupayaan pengiraan, kaedah terkini iaitu kaedah kebolehjadian maksimum merupakan antara cara yang terbaik untuk menguruskan masalah data lenyap. Di dalam kertas ini, dua kaedah gantian moden diperkenalkan iaitu jangkaan pemaksimuman (EM) dan jangkaan pemaksimum bootstrap (EMB) untuk digunakan di dalam model linear hubungan fungsian (LFRM) iaitu

show abstract

“…The EM algorithm has two steps which are applied alternately in an iterative fashion. Each step is guaranteed to increase the likelihood of the observed data, and the algorithm converges to a local maximum of the likelihood function [12,15]. The method gives for each component and each observation a probability of the observation stemming from that component.…”

Section: Probabilistic Modelingmentioning

confidence: 99%

Rule Discovery and Probabilistic Modeling for Onomastic Data

Leino

Mannila

Pitkänen

2003

Knowledge Discovery in Databases: PKDD 2003

View full text Add to dashboard Cite

Abstract. The naming of natural features, such as hills, lakes, springs, meadows etc., provides a wealth of linguistic information; the study of the names and naming systems is called onomastics. We consider a data set containing all names and locations of about 58,000 lakes in Finland. Using computational techniques, we address two major onomastic themes. First, we address the existence of local dependencies or repulsion between occurrences of names. For this, we derive a simple form of spatial association rules. The results partially validate and partially contradict results obtained by traditional onomastic techniques. Second, we consider the existence of relatively homogeneous spatial regions with respect to the distributions of place names. Using mixture modeling, we conduct a global analysis of the data set. The clusterings of regions are spatially connected, and correspond quite well with the results obtained by other techniques; there are, however, interesting differences with previous hypotheses.

show abstract

On the Convergence Properties of the EM Algorithm

Cited by 2,691 publications

References 17 publications

Parametric fractional imputation for missing data analysis

Parametric fractional imputation for missing data analysis

Missing Value Estimation Methods for Data in Linear Functional Relationship Model

Rule Discovery and Probabilistic Modeling for Onomastic Data

Contact Info

Product

Resources

About