Kd-tree and adaptive radius (KD-AR Stream) based real-time data stream clustering

Creative Commons License

Senol A., Karacan H.

JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY, cilt.35, sa.1, ss.337-354, 2020 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 35 Sayı: 1
  • Basım Tarihi: 2020
  • Doi Numarası: 10.17341/gazimmfd.467226
  • Dergi Adı: JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Art Source, Compendex, TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.337-354
  • Anahtar Kelimeler: Data stream clustering, kd-tree, adaptive radius, fully online, evolving clustering, SENSOR NETWORKS, ALGORITHM, SEARCH, FUSION
  • Gazi Üniversitesi Adresli: Evet

Özet

Data stream clustering is one of the most popular topics of today's world where the amount of data reaches incredible levels in parallel with technological developments. The most important problems encountered in data stream clustering approaches are the fact that most of the approaches consists of an online and offline phases, the definition of the number of cluster, or the need to set a limitation to this number, the problems encountered in determining optimum radius value, and the problems encountered in concept evolution. The present study proposes an evolutionary based solution method, which is based on Kd-Tree and adaptive radius (KD-AR Stream) to perform real-time clustering on the streaming data. The proposed approach has been compared with SE-Stream, DPStream and CEDAS algorithms in terms of both cluster quality and execution time. The results showed that KD-AR Stream algorithm has a good clustering performance within a reasonable time by comparison with the other algorithms.