arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1803.04579cs.LGcs.CL
收藏

也是训练数据修剪!

It was the training data pruning too!

The University of Chicago(芝加哥大学);Google Research(谷歌研究院)

Pramod Kaushik Mudrakarta, Ankur Taly, Mukund Sundararajan, Kedar Dhamdhere

摘要:

我们研究了当前最好的模型(KDG)的问答的表格数据评估WikiTableQuestions数据集。先前针对该模型进行的消融研究将该模型的性能归因于其架构的某些方面。在本文中,我们发现,该模型的性能也很大程度上取决于一定的修剪数据用于训练模型。禁用修剪步骤会使模型的准确率从43.3%下降到36.3%。对KDG模型性能的巨大影响表明,修剪也可能是训练其他语义解析器的有用预处理步骤。

英文摘要:

We study the current best model (KDG) for question answering on tabular data evaluated over the WikiTableQuestions dataset. Previous ablation studies performed against this model attributed the model's performance to certain aspects of its architecture. In this paper, we find that the model's performance also crucially depends on a certain pruning of the data used to train the model. Disabling the pruning step drops the accuracy of the model from 43.3% to 36.3%. The large impact on the performance of the KDG model suggests that the pruning may be a useful pre-processing step in training other semantic parsers as well.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑