Optimal Transport based Drift Detection for Sensor Streams: Method and Applications in Transportation

A. K. Laha and Shikha Verma

IIMA Working Papers from Indian Institute of Management Ahmedabad, Research and Publication Department

Abstract: With increasing adoption of Internet of Things (IoT) across the transportation sector, there is a growing need for developing algorithms for analyzing data streams. Due to dynamic operating environment conditions in the transportation domain, the nature of the data streams frequently change and static predictive models are often not successful when dealing with, non-stationary data streams. Further, labelled data is often unavailable or is costly to acquire in real time. Thus, effective algorithms for such problems would aim to maximize accuracy while minimizing the labelled data requirements. In this paper, we propose a new algorithm namely, the Optimal Transport based Drift Detection (OTDD) algorithm, that aims to address the accuracy-labeling requirement trade-off. Experiments on artificial and real-life data sets from the transportation domain demonstrate that the OTDD algorithm performs better than some of the widely used competing algorithms in addressing the accuracy-labeling requirement trade-off.

Date: 2021-09-09
New Economics Papers: this item is included in nep-ict, nep-isf and nep-tre
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.iima.ac.in/sites/default/files/rnpfiles/16464727852021-09-01.pdf English Version (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:iim:iimawp:14660

Access Statistics for this paper

More papers in IIMA Working Papers from Indian Institute of Management Ahmedabad, Research and Publication Department Contact information at EDIRC.
Bibliographic data for series maintained by (respub@iima.ac.in).