Learning Dynamic Feature Selection for Fast Sequential Prediction

@article{Strubell2015LearningDF,
  title={Learning Dynamic Feature Selection for Fast Sequential Prediction},
  author={Emma Strubell and L. Vilnis and Kate Silverstein and A. McCallum},
  journal={ArXiv},
  year={2015},
  volume={abs/1505.06169}
}
We present paired learning and inference algorithms for significantly reducing computation and increasing speed of the vector dot products in the classifiers that are at the heart of many NLP components. This is accomplished by partitioning the features into a sequence of templates which are ordered such that high confidence can often be reached using only a small fraction of all features. Parameter estimation is arranged to maximize accuracy and early confidence in this sequence. Our approach… Expand
13 Citations
Speed-Accuracy Tradeoffs in Tagging with Variable-Order CRFs and Structured Sparsity
  • 8
  • PDF
Dynamic Feature Induction: The Last Gist to the State-of-the-Art
  • 45
  • PDF
Learning to Prune: Exploring the Frontier of Fast and Accurate Parsing
  • 10
  • PDF
Resource Constrained Structured Prediction
  • 8
  • PDF
An Algebra for Feature Extraction
  • 2
  • PDF
Building a Robust Text Classifier on a Test-Time Budget
  • PDF
Test time feature ordering with FOCUS: interactive predictions with minimal user burden
  • 14
  • PDF
Reward Augmented Maximum Likelihood for Neural Structured Prediction
  • 152
  • PDF
...
1
2
...

References

SHOWING 1-10 OF 32 REFERENCES
Fast and Robust Part-of-Speech Tagging Using Dynamic Model Selection
  • 31
  • PDF
Vine Pruning for Efficient Multi-Pass Dependency Parsing
  • 58
  • PDF
Dynamic Feature Selection for Dependency Parsing
  • 49
  • PDF
Structured Prediction Cascades
  • 115
  • PDF
Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network
  • 3,254
  • PDF
Learning Adaptive Value of Information for Structured Prediction
  • 31
  • PDF
Structured Sparsity in Structured Prediction
  • 74
  • Highly Influential
  • PDF
Classifier Cascade for Minimizing Feature Evaluation Cost
  • 68
  • PDF
Support vector machine learning for interdependent and structured output spaces
  • 1,429
  • PDF
Search-based structured prediction
  • 529
  • PDF
...
1
2
3
4
...