Time-Series Big Data Stream Evaluation

Petrus Mursanto, Ari Wibisono, Wendy D.W.T. Bayu, Valian Fil Ahli, May Iffah Rizki, Lintang Matahari Hasani, Jihan Adibah

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Big data processing is a challenging job. Extensive time-series data need a method of preparation, management, and feature calculation for each data arrival. FIMT-DD is an algorithm for processing predictive regression for big data. The splitting criteria in the standard FIMT-DD algorithm use a Hoeffding Bound. We propose to change the splitting criteria to Chernoff bound. The experimental results and the performance comparisons that we did have better results than the standard method. We use three real-world datasets. The improvement that we propose can produce a 2.3% accuracy improvement for traffic demand data.

Original languageEnglish
Title of host publication2020 International Workshop on Big Data and Information Security, IWBIS 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages41-45
Number of pages5
ISBN (Electronic)9781728190983
DOIs
Publication statusPublished - 17 Oct 2020
Event5th International Workshop on Big Data and Information Security, IWBIS 2020 - Depok, Indonesia
Duration: 17 Oct 202018 Oct 2020

Publication series

Name2020 International Workshop on Big Data and Information Security, IWBIS 2020

Conference

Conference5th International Workshop on Big Data and Information Security, IWBIS 2020
CountryIndonesia
CityDepok
Period17/10/2018/10/20

Keywords

  • Big Data
  • Chernoff Bound
  • Data Stream
  • FIMT-DD
  • Intelligent Systems
  • Standard Deviation

Fingerprint Dive into the research topics of 'Time-Series Big Data Stream Evaluation'. Together they form a unique fingerprint.

Cite this