|
Lehrende/r: Zahra Ahmadi
Veranstaltungsart:
online: Vorlesung/Übung
Anzeige im Stundenplan:
08.079.20510
Semesterwochenstunden:
4
Credits:
6,0
Unterrichtssprache:
Englisch
Min. | Max. Teilnehmerzahl:
- | -
Inhalt:
- introduction to data streams, basics of streams: windowing, sliding-window
- DSMS, prototype systems (Aurora, STREAM, TelegraphCQ), data stream warehousing (and OLAP), load shedding
- synopsis construction (sampling, wavelet, sketches, histograms, Bloom filter, random projections, LSH, PCA, ...)
- change detection methods (CUSUM, SPC, martingales, ...)
- outlier detection/novelty detection
- classification of data streams (ensemble classifiers {Littlestone,WINNOW,...}, VFDT, CVFDT, ...)
- graph stream mining (sketches, link prediction, ...)
- distributed stream mining (using Flink)
Zusätzliche Informationen:
- Data Streams Models and Algorithms, Charu C. Aggarwal, Springer
- Knowledge Discovery from Data Streams, João Gama, CRC
- Data Stream Management, Lukasz Golab & M. Tamer Özsu, Morgan & Claypool publishers
- Data Streams: Algorithms and Applications, S. Muthukrishnan, now
- Graph Stream Algorithms: A Survey, Andrew McGregor, SIGMOD 2014
- Mining of Massive Datasets, Jure Leskovec & Anand Rajaraman & Jeffrey D. Ullman, ch.4
Zugeordnete Lehrveranstaltungen:
Diese Veranstaltung kann nicht im Master im Spezialisierungsbereich eingebracht werden.
|
