Access the full text.
Sign up today, get DeepDyve free for 14 days.
Chapter 4 4.1 The Likelihood of Finite Mixture Models Estimation of the parameters of the mixing distribution P is predominantly done using maximum likelihood. Given a sample of iid x ∼ f (x|P), i = 1,..., n, (4.1) we are interested in ﬁnding the maximum likelihood estimates (MLEs) of P, denoted as P, that is P = arg max L(P), n k L(P)= f (x , λ ) p (4.2) i j j ∏ ∑ i=1 j=1 or alternatively ﬁnding the estimates of P which maximize the log likelihood function n k (P)= log L(P)= log f (x , λ ) p . (4.3) i j j ∑ ∑ i=1 j=1 An estimate of P can be obtained as a solution to the likelihood equation ∂ (P) S(x, P)= = 0, (4.4) ∂ P where S(x, P) is the gradient vector of the log likelihood function, where differenti- ation is with respect to the parameter vector P. Maximum likelihood estimation of P is by no means trivial, since there are mostly no closed-form solutions available. P. Schlattmann, Medical Applications of Finite Mixture Models,55 Statistics for Biology and Health, DOI: 10.1007/978-3-540-68651-4 4, Springer-Verlag Berlin Hiedelberg 2009 56 4 Theory and
Published: Dec 8, 2008
Keywords: Mixture Model; Convex Hull; Bayesian Information Criterion; Expectation Maximization; Directional Derivative
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.