Current derivative estimation of non-stationary processes based on metrical information

Elena Kochegurova, Ekaterina Gorokhova

Research output: Contribution to journalConference articlepeer-review

5 Citations (Scopus)

Abstract

Demand for estimation of derivatives has arisen in a range of some applied problems. One of the possible approaches to estimating derivatives is to approximate measurement data. The problem of real-time estimation of derivatives is investigated. A variation method of obtaining recurrent smoothing splines is proposed for estimation of derivatives. A distinguishing feature of the described method is recurrence of spline coefficients with respect to its segments and locality about measured values inside the segment. Influence of smoothing spline parameters on efficiency of such estimations is studied. Comparative analysis of experimental results is performed.

Original languageEnglish
Pages (from-to)512-519
Number of pages8
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9330 LNCS
DOIs
Publication statusPublished - 2015
Event7th International Conference on Computational Collective Intelligence, ICCCI 2015 - Madrid, Spain
Duration: 21 Sep 201523 Sep 2015

Keywords

  • Derivatives estimation
  • Recurrent algorithm
  • Variation smoothing spline

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Current derivative estimation of non-stationary processes based on metrical information'. Together they form a unique fingerprint.

Cite this