JavaScript Implementation of Scagnostics and Its Applications

Author:

Pham Vung,Dang Tommy

Abstract

Scagnostics is a set of features that characterizes the 2D distributions in the underlying data. Various real-world applications have been using Scagnostics visual features to detect unusual bivariate data correlations. Concomitantly, many applications are required to be implemented on web platforms due to their accessibility and convenience. Therefore, this chapter discusses a recent JavaScript implementation of Scagnostics, an extension to higher dimensional data, and its applications in detecting abnormalities in bivariate and multivariate time series data. Its implementation in JavaScript supports the tremendous demand for visual features in the web environment. Likewise, its higher dimensional implementations allow generating Scagnostics features for the rapidly growing multivariate data. Finally, conventional ScagnosticsJS computations involve time-consuming algorithms, and they are sensitive to slight changes in the underlying data. Therefore, this chapter also discusses a recent attempt to tackle these issues using machine learning to estimate the Scagnostics scores.

Publisher

IntechOpen

Reference26 articles.

1. Wilkinson L, Anand A, Grossman R. Graph-theoretic scagnostics. Proceedings - IEEE Symposium on Information Visualization, INFO VIS. 2005:157–164.

2. Lee W, Anushka A. Compute scagnostics - scatterplot diagnostics. rforge; 2018. Available from: https://www.rforge.net/scagnostics/.

3. Josua K. Python binding to R scagnostics. github; 2015. Available from: https://github.com/nyuvis/scagnostics.

4. Dang TN, Wilkinson L. ScagExplorer: Exploring scatterplots by their scagnostics. IEEE Pacific Visualization Symposium. 2014:73–80.

5. Fu L. Implementation of Three-dimensional Scagnostics [master’s thesis]. University of Waterloo; 2009.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3