An O(n log n) Cutting Plane Algorithm for Structured Output Ranking

Abstract : In this work, we consider ranking as a training strategy for structured output prediction. Recent work has begun to explore structured output prediction in the ranking setting, but has mostly focused on the special case of bipartite preference graphs. The bipartite special case is computationally efficient as there exists a linear time cutting plane training strategy for hinge loss bounded regularized risk, but it is unclear how to feasibly extend the approach to complete preference graphs. We develop here a highly parallelizable O(n log n) algorithm for cutting plane training with complete preference graphs that is scalable to millions of samples on a single core. We explore theoretically and empirically the relationship between slack rescaling and margin rescaling variants of the hinge loss bound to structured losses, showing that the slack rescaling variant has better stability properties and empirical performance with no additional computational cost per cutting plane iteration. We further show generalization bounds based on uniform convergence. Finally, we demonstrate the effectiveness of the proposed family of approaches on the problem of object detection in computer vision.
Complete list of metadatas

Cited literature [21 references]  Display  Hide  Download

https://hal.inria.fr/hal-01020943
Contributor : Matthew Blaschko <>
Submitted on : Tuesday, July 15, 2014 - 4:54:44 PM
Last modification on : Thursday, February 7, 2019 - 5:29:19 PM
Long-term archiving on : Friday, October 16, 2015 - 2:47:16 PM

File

StructuredOutputRanking.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Matthew Blaschko, Arpit Mittal, Esa Rahtu. An O(n log n) Cutting Plane Algorithm for Structured Output Ranking. German Conference on Pattern Recognition, Sep 2014, Münster, Germany. ⟨10.1007/978-3-319-11752-2_11⟩. ⟨hal-01020943⟩

Share

Metrics

Record views

500

Files downloads

650