Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Cited by Google
- Similars in SciELO
- Similars in Google
Share
Revista de Ciencias
Print version ISSN 0121-1935
Abstract
ARENAS, Favián; MARTINEZ, Héctor Jairo and PEREZ, Rosana. Structured BFGS Method for the Estimation of the Maximum Likelihood. rev. cienc. [online]. 2016, vol.20, n.2, pp.39-54. ISSN 0121-1935. https://doi.org/10.25100/rc.v20i2.4672.
Given the special structure of the Hessian matrix of the log-likelihood function which is parallel to that found in nonlinear least-squares problems, we introduce the structured BFGS secant method for the maximum-likelihood estimation and for the development or the local and super-linear convergence theory for the algorithm, following the lines of Martínez; Martínez and Egels, and the theory about maximum-likelihood estimation given in Gonglewski. We present the results of some numerical experiments which show a good performance of our algorithm.
Keywords : maximum likelihood estimation; likelihood function; structured secant method; nonlinear least‐squares problem; super-linear convergence.