×
Sep 22, 2020 · This paper covers two main points: (i) we propose a faster FDA alternative that benefits from parallelization and (ii) improve the performance of FDA.
Feature Decay Algorithms (FDA) are a technique for data selection that has demonstrated excellent performance in a number of tasks. This method maximizes the ...
Feature Decay Algorithms (FDA) are a technique for data selection that has demonstrated excellent performance in a number of tasks. This method maximizes the ...
In this work, we use Feature Decay Al- gorithms (FDA), a transductive data selection method that has achieved good results in SMT and apply it to NMT. Our ...
Nov 7, 2018 · In this paper we propose to use an approximated target-side in addition to the source-side when selecting suitable sentence-pairs for training a model.
We use feature decay algorithms (FDA) for fast deployment of accurate statis- tical machine translation systems taking only about half a day for each transla-.
We use feature decay algorithms (FDA) for fast deployment of accurate statistical machine translation systems taking only about half a day for each translation ...
People also ask
The FDA algorithm has already demonstrated its competitiveness by achieving excellent results in several Workshops on SMT from 2013–2015, on both MT and quality ...
Improved feature decay algorithms for statistical machine translation · Computer Science. Natural Language Engineering · 2020.
Feature decay algorithms for fast deployment of accurate statistical machine translation systems. 2013. Biçici, Ergun and Deniz Yuret. Instance selection ...