Semistochastic Quadratic Bound Methods
This work addresses inference challenges in machine learning models such as conditional random fields and logistic regression, offering an incremental improvement over existing batch and stochastic methods.
The paper tackles the problem of maximum likelihood inference for partition functions in models like conditional random fields and logistic regression by proposing semistochastic quadratic bound (SQB) methods, which achieve global convergence and linear convergence rates under certain assumptions, and demonstrate efficacy through comparisons with state-of-the-art techniques on common datasets.
Partition functions arise in a variety of settings, including conditional random fields, logistic regression, and latent gaussian models. In this paper, we consider semistochastic quadratic bound (SQB) methods for maximum likelihood inference based on partition function optimization. Batch methods based on the quadratic bound were recently proposed for this class of problems, and performed favorably in comparison to state-of-the-art techniques. Semistochastic methods fall in between batch algorithms, which use all the data, and stochastic gradient type methods, which use small random selections at each iteration. We build semistochastic quadratic bound-based methods, and prove both global convergence (to a stationary point) under very weak assumptions, and linear convergence rate under stronger assumptions on the objective. To make the proposed methods faster and more stable, we consider inexact subproblem minimization and batch-size selection schemes. The efficacy of SQB methods is demonstrated via comparison with several state-of-the-art techniques on commonly used datasets.