Browsing Tag
disjoint pseudolikelihood
1 post
Definition
Disjoint pseudolikelihood is a computational approximation method for the Ising model. The normalization constants for DPL consist of p terms rather than 2^p terms in the exact likelihood, making it computationally less demanding. However, maximum pseudolikelihood estimation based on DPL has different performance characteristics compared to other estimation approaches.
Sources: Keetelaar et al. (2024)
Related Terms
Applications
Disjoint Pseudolikelihood and Joint Pseudolikelihood
Disjoint pseudolikelihood and joint pseudolikelihood are two competing approximations for the Ising model.
Sources: Keetelaar et al. (2024)
Disjoint Pseudolikelihood and the Ising Model
Disjoint pseudolikelihood serves as a computationally tractable alternative for parameter estimation in the Ising model.
Sources: Keetelaar et al. (2024)
Research Articles
Comparing maximum likelihood and maximum pseudolikelihood estimators for the Ising model
By Sara Keetelaar, Nikola Sekulovski, Denny Borsboom, & Maarten Marsman