Computer Science
8 years ago

EM Algorithm Revolutionizes Maximum Likelihood Estimation for Faster Convergence Rates

Journal of the Royal Statistical Society Series B (Methodological)
On Use of the Em Algorithm for Penalized Likelihood Estimation
Peter J. Green

Paper Summary

The EM algorithm is commonly used for estimating likelihood, but not often for penalized likelihood. This paper explores how the EM algorithm performs in penalized likelihood estimation, focusing on how quickly it converges. The researchers found an alternative method that is usually more practical and converges just as fast.

Related papers

Theme background image
Mathematics
3 years ago

New algorithm revolutionizes solving nonlinear equations with missing data.

Missing Data
Soil Mapping
Multicollinearity
2 views
Theme background image
Mathematics
5 years ago

New method for precise estimation and testing in data analysis

Multicollinearity
Regularization Methods
Missing Data
2 views
Theme background image
Mathematics
8 years ago

EM algorithm improves accuracy of K-distribution parameter estimation for large datasets.

Skew Distributions
Missing Data
Regularization Methods
2 views