Parameter Learning for Loopy Markov Random Fields with Structural Support Vector Machines
نویسندگان
چکیده
Discriminative learners for functions with structured outputs have been successfully applied to sequence prediction, parsing, sequence alignment, and other problems where exact inference is tractable. However, these learners face challenges when inference procedures must use approximations, e.g., for multi-label classification, clustering, image segmentation, or loopy graphical models. In this paper, we explore methods for training structural SVMs on problems where exact inference is intractable. In particular, we consider pairwise fully connected Markov random fields, using multi-label classification as an example application. We show how to adapt loopy belief propagation, greedy search, and linear programming relaxations as (approximate) separation oracles in the structural SVM cutting-plane training algorithm. In addition to theoretical characterizations, we empirically evaluate and analyze the resulting algorithms on six multi-label classification datasets.
منابع مشابه
A Comparative Study of Extreme Learning Machines and Support Vector Machines in Prediction of Sediment Transport in Open Channels
The limiting velocity in open channels to prevent long-term sedimentation is predicted in this paper using a powerful soft computing technique known as Extreme Learning Machines (ELM). The ELM is a single Layer Feed-forward Neural Network (SLFNN) with a high level of training speed. The dimensionless parameter of limiting velocity which is known as the densimetric Froude number (Fr) is predicte...
متن کاملConstrained Approximate Maximum Entropy Learning of Markov Random Fields
Parameter estimation in Markov random fields (MRFs) is a difficult task, in which inference over the network is run in the inner loop of a gradient descent procedure. Replacing exact inference with approximate methods such as loopy belief propagation (LBP) can suffer from poor convergence. In this paper, we provide a different approach for combining MRF learning and Bethe approximation. We cons...
متن کاملMining Biological Repetitive Sequences Using Support Vector Machines and Fuzzy SVM
Structural repetitive subsequences are most important portion of biological sequences, which play crucial roles on corresponding sequence’s fold and functionality. Biggest class of the repetitive subsequences is “Transposable Elements” which has its own sub-classes upon contexts’ structures. Many researches have been performed to criticality determine the structure and function of repetitiv...
متن کامل248 Remotely Sensed Data Characterization
EMPs Extended morphological profiles EMPs Extended morphological profiles LDA Linear discriminant analysis LogDA Logarithmic discriminant analysis MLR Multinomial logistic regression MLRsubMRF Subspace-based multinomial logistic regression followed by Markov random fields MPs Morphological profiles MRFs Markov random fields PCA Principal component analysis QDA Quadratic discriminant analysis RH...
متن کاملBayesian Learning in Undirected Graphical Models: Approximate MCMC Algorithms
Bayesian learning in undirected graphical models—computing posterior distributions over parameters and predictive quantities— is exceptionally difficult. We conjecture that for general undirected models, there are no tractable MCMC (Markov Chain Monte Carlo) schemes giving the correct equilibrium distribution over parameters. While this intractability, due to the partition function, is familiar...
متن کامل