Partial Inference in Structured Prediction
Preprint
Abstract
In this paper, we examine the problem of partial inference in the context of structured prediction. Using a generative model approach, we consider the task of maximizing a score function with unary and pairwise potentials in the space of labels on graphs. Employing a two-stage convex optimization algorithm for label recovery, we analyze the conditions under which a majority of the labels can be recovered. We introduce a novel perspective on the Karush-Kuhn-Tucker (KKT) conditions and primal and dual construction, and provide statistical and topological requirements for partial recovery with provable guarantees.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- Exact Recovery in the Stochastic Block ModelIEEE Transactions on Information Theory, 2015
- An Introduction to Matrix Concentration InequalitiesFoundations and Trends in Machine Learning, 2015
- Structured Learning and Prediction in Computer VisionFoundations and Trends in Computer Graphics and Vision, 2011
- High-dimensional covariance estimation by minimizing ℓ1-penalized log-determinant divergenceElectronic Journal of Statistics, 2011