Current Machine Translation systems achieve very good results on a growing
variety of language pairs and data sets. However, it is now well known that
they produce fluent translation outputs that often can contain important
meaning errors. Quality Estimation task deals with the estimation of quality of
translations produced by a Machine Translation system without depending on
Reference Translations. A number of approaches have been suggested over the
years. In this paper we show that the parallel corpus used as training data for
training the MT system holds direct clues for estimating the quality of
translations produced by the MT system. Our experiments show that this simple
and direct method holds promise for quality estimation of translations produced
by any purely data driven machine translation system.

本文研究了使用平行语料库来估计由数据驱动机器翻译系统产生的翻译质量，并展示了该简单直接的方法对机器翻译系统的翻译质量估计具有潜在的应用价值。

基于训练数据直接证据的机器翻译文本质量评估

Quality Estimation of Machine Translated Texts based on Direct Evidence  from Training Data

We investigate data driven localized wave solutions of the Fokas-Lenells
equation by using physics informed neural network(PINN). We improve basic PINN
by incorporating control parameters into the residual loss function. We also
add conserve quantity as another loss term to modify the PINN. Using modified
PINN we obtain the data driven bright soliton and dark soliton solutions of
Fokas-Lenells equation. Conserved quantities informed loss function achieve
more accuracy in terms of relative L2 error between predicted and exact soliton
solutions. We hope that the present investigation would be useful to study the
applications of deep learning in nonlinear optics and other branches of
nonlinear physics. Source codes are available at
this https URL

使用物理学中的神经网络来研究数据驱动的局部波动解法，为非线性光学和其他非线性物理学领域应用深度学习提供可能。