A Learning Error Analysis for Structured Prediction with Approximate Inference

Authors: Yuanbin Wu, Man Lan, Shiliang Sun, Qi Zhang, Xuanjing Huang

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical evaluations on text classification, sequential labelling and dependency parsing witness the success of approximate inference and the benefit of the proposed margin.
Researcher Affiliation Academia 1School of Computer Science and Software Engineering, East China Normal University 2Shanghai Key Laboratory of Multidimensional Information Processing 3School of Computer Science, Fudan University
Pseudocode Yes Algorithm 3: Structured perceptron with mρ. and Algorithm 4: Online PA with mρ.
Open Source Code No The paper references third-party tools like SVMmulticlass and MSTParser, but does not provide a link or statement about the authors' own open-source code for the described methodology.
Open Datasets Yes We report results on the 20 newsgroups corpus (18000 documents, 20 classes). We use the Co NLL 2000 dataset [Sang and Buchholz, 2000] for chunking and POS tagging, SIGHAN 2005 bake-off corpus (pku and msr) [Emerson, 2005] for word segmentation.
Dataset Splits Yes We take 20% of the training set as development set for tuning ρ (grid search in [0, 2] with step size 0.05).
Hardware Specification No The paper does not provide any specific details about the hardware used for the experiments.
Software Dependencies Yes The implementation is adapted from SVMmulticlass 3.
Experiment Setup Yes We take 20% of the training set as development set for tuning ρ (grid search in [0, 2] with step size 0.05). We use Algorithm 3 with 20 iterations and learning step 1.