Experts

1

2

3

4

5

6

7

8

9

10
Publications (18)

1Entropybased Training Data Selection for Domain AdaptationMoore and Lewis (2010) calculated the difference of the cross entropy values for a given sentence, based on language models...

2From FiniteState to Inversion Transductions: Toward Unsupervised Bilingual Grammar Inductiongrammars. We show a consistent improvement in terms of crossentropy throughout the bootstrapping process, as well as promising...

3KullbackLeibler Distance between Probabilistic ContextFree Grammars and Probabilistic Finite Automata2004  MarkJan Nederhof,Giorgio Sattafor one part of the KullbackLeibler distance, viz. the crossentropy. We discuss several applications of the result to the...

4SyntaxDriven Machine Translation as a Model of ESL Revision2010  Huichao Xue,Rebecca Hwaexpected structure. 3.3 Method of Model Comparison Cross entropy can be used as a metric that measures the distance between...

5A Stochastic Parser Based on a Structural Word Prediction Modelthe predictive power of our model, we calculated their cross entropy on the test corpns. In this process, the annotated tree...

6Language Modeling for Spoken Dialogue System based on Filtering using PredicateArgument Structuresdocuments. The relevance measure can be defined with crossentropy or perplexity by the language model generated from the...

7Translation QualityBased Supplementary Data Selection by Incremental Update of Translation Modelstechniques (Hildebrand et al., 2005) to perplexity or crossentropy on ‘indomain’ datasets (Foster and Kuhn, 2007; Banerjee...

8Probabilistic Refinement Algorithms for the Generation of Referring Expressionscolumns were assigned the same probability. Figure 4: Crossentropy between the corpus distribution and different runs of...

9Committeebased Decision Making in Probabiiistic Partial Parsing2000  Takashi Inui,Kentaro Inuialso be seen as an effort for reducing the averaged cross entropy of the model on test, data. Since PA curves tend to...

10A Comparative Study on Ranking and Selection Strategies for MultiDocument Summarizationfunction. The objective of the algorithm is to minimize the cross entropy between the gold standard probability and the predicted...