arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1301.4171cs.IRcs.LGstat.ML
收藏

仿射加权嵌入

Affinity Weighted Embedding

Google Inc.(谷歌公司)

Jason Weston, Ron Weiss, Hector Yee

摘要:

像Wsabie和PSI这样的监督(线性)嵌入模型已经证明在排名,推荐和注释任务中是成功的。然而,尽管可扩展到大型数据集,但由于其线性性质,它们并没有充分利用额外的数据,并且通常欠拟合。我们提出了一类新的模型,其目的是提供更好的性能,同时保留许多现有的嵌入模型类的好处。我们的新方法通过迭代学习线性嵌入模型来工作,其中下一次迭代的特征和标签作为上一次迭代的函数进行重新加权。我们描述了几个变种的家庭,并给出了一些初步的结果。

英文摘要:

Supervised (linear) embedding models like Wsabie and PSI have proven successful at ranking, recommendation and annotation tasks. However, despite being scalable to large datasets they do not take full advantage of the extra data due to their linear nature, and typically underfit. We propose a new class of models which aim to provide improved performance while retaining many of the benefits of the existing class of embedding models. Our new approach works by iteratively learning a linear embedding model where the next iteration's features and labels are reweighted as a function of the previous iteration. We describe several variants of the family, and give some initial results.

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。

↑