arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2111.12795cs.HCcs.AIcs.LG
收藏

Picasso:无模型特征可视化

Picasso: Model-free Feature Visualization

University of Southern California(南加州大学);Meta

Binh Vu, Igor Markov

摘要:

如今,机器学习(ML)应用程序可以访问成千上万的功能。有了这样的特征集,有效地浏览和管理最相关特征的子集是一个挑战。在本文中,我们提出了一种新的方法来可视化多达数千个功能在一个单一的图像。图像不仅显示了单个特征的信息,而且还通过特征的相对定位来表达特征的相互作用。

英文摘要:

Today, Machine Learning (ML) applications can have access to tens of thousands of features. With such feature sets, efficiently browsing and curating subsets of most relevant features is a challenge. In this paper, we present a novel approach to visualize up to several thousands of features in a single image. The image not only shows information on individual features, but also expresses feature interactions via the relative positioning of features.

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。

↑