Learning to Prune: Exploring the Frontier of Fast and Accurate Parsing
Tim Vieira,Jason Eisner +1 more
Reads0
Chats0
TLDR
The authors train a pruning policy under an objective that measures end-to-end performance: they search for a fast and accurate policy, which leads to a better Pareto frontier.Abstract:
Pruning hypotheses during dynamic programming is commonly used to speed up inference in settings such as parsing. Unlike prior work, we train a pruning policy under an objective that measures end-to-end performance: we search for a fast and accurate policy. This poses a difficult machine learning problem, which we tackle with the LOLS algorithm. LOLS training must continually compute the effects of changing pruning decisions: we show how to make this efficient in the constituency parsing setting, via dynamic programming and change propagation algorithms. We find that optimizing end-to-end performance in this way leads to a better Pareto frontier---i.e., parsers which are more accurate for a given runtime.read more
Citations
More filters
Proceedings ArticleDOI
Unsupervised Parsing with S-DIORA: Single Tree Encoding for Deep Inside-Outside Recursive Autoencoders
TL;DR: S-DIORA, an improved variant of DIORA that encodes a single tree rather than a softly-weighted mixture of trees by employing a hard argmax operation and a beam at each cell in the chart, is introduced.
Proceedings Article
Generic refinement of expressive grammar formalisms with an application to discontinuous constituent parsing.
TL;DR: This work forms a generalization of Petrov et al. (2006)’s split/merge algorithm for interpreted regular tree grammars (Koller and Kuhlmann, 2011) that captures a large class of grammar formalisms.
Journal ArticleDOI
Distilled Neural Networks for Efficient Learning to Rank
TL;DR:
Posted Content
Efficient Computation of Expectations under Spanning Tree Distributions
TL;DR: A general framework for inference in spanning tree models is given and unified algorithms for the important cases of first- order expectations and second-order expectations in edge-factored, non-projective spanning-tree models are proposed.
Journal ArticleDOI
Machine Learning in Real-Time Internet of Things (IoT) Systems: A Survey
Jiang Bian,Abdullah Arafat,Haoyi Xiong,Jing Li,Li Li,Hongyang Chen,Jun Wang,Dejing Dou,Zhishan Guo +8 more
TL;DR: This article will rigorously explore the state-of-the-art results emphasizing the strengths and weaknesses in ML/DL-based scheduling techniques, accuracy versus execution time tradeoff policies of ML algorithms, and security and privacy of learning-based algorithms in real-time IoT systems.
References
More filters
ReportDOI
Building a large annotated corpus of English: the penn treebank
TL;DR: As a result of this grant, the researchers have now published on CDROM a corpus of over 4 million words of running text annotated with part-of- speech (POS) tags, which includes a fully hand-parsed version of the classic Brown corpus.
Proceedings ArticleDOI
Minimum Error Rate Training in Statistical Machine Translation
TL;DR: It is shown that significantly better results can often be obtained if the final evaluation criterion is taken directly into account as part of the training procedure.
Proceedings ArticleDOI
A Fast and Accurate Dependency Parser using Neural Networks
TL;DR: This work proposes a novel way of learning a neural network classifier for use in a greedy, transition-based dependency parser that can work very fast, while achieving an about 2% improvement in unlabeled and labeled attachment scores on both English and Chinese datasets.
Proceedings Article
A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning
TL;DR: In this article, a no-regret algorithm is proposed to find a policy with good performance under the distribution of observations it induces in such sequential settings, which can be seen as a no regret algorithm in an online learning setting.
Posted Content
A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning
TL;DR: In this article, a no-regret algorithm is proposed to train a stationary deterministic policy with good performance under the distribution of observations it induces in such sequential settings, and it outperforms previous approaches on two challenging imitation learning problems and a benchmark sequence labeling problem.