An Algorithm for Streaming Clustering
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
A simple existing data stream clustering algorithm DenStream based on DBScan is studied. Based on DenStream a modified algorithm called DenStream2 is proposed. It follows most of the framework and theory of DenStream. Denstream2 is implemented as a foreign function in an extensible data stream management system (DSMS), where queries over streams are allowed. The generated clusters inferred from each window of an input a data stream are emitted as new stream clusters. The output stream can be stored in database for later queries, or be queried directly.
Keywords: DBScan, DenStream
Place, publisher, year, edition, pages
IT, 11 011
IdentifiersURN: urn:nbn:se:uu:diva-150820OAI: oai:DiVA.org:uu-150820DiVA: diva2:408944
Risch, ToreJansson, Anders