Data flow mining and anomaly detection

Presented are data streams, tasks and challenges of mining the data stream. The emphasis is on practical work, which presents the detection of abnormalities, a special form of data mining. Participants get familiar with storing the data stream, which is vital for data mining and anomaly detection.

Content:

  • Data flows
    • Model of the data stream, the sampling and filtering of the data in the stream, counting the various elements in the stream, evaluating events, counting the elements in the interval
  • Anomaly Detection
    • What are the anomalies, challenges/tasks of detection of anomalies, types of anomalies and anomaly detection methods, applications for detecting anomalies
  • Cassandra presentation and benefits
  • Example of processing of the data stream
  • Interval storage of data in Cassandra
  • Implementation of counting the elements in the interval
  • Detection of sudden jumps occurrence of the event as an example of detection of anomalies

Target group

Načrtovalci IoT in Big Data storitev
R&D specialisti

Goals

The purpose of the workshop is to know the data streams and methods for extracting information from them, know the anomalies and their detection methods. Participants become familiar with the process of data mining data stream and detecting anomalies on real problems.
Difficulty:
Type:LectureDemonstration
Status:Not exceeded
Starts:Wednesday, 10. October 2018, 08:30am
Ends:Wednesday, 10. October 2018, 15:30pm
Duration:1 day
Tutor:Matej Kren
Location: Fakulteta za elektrotehniko
Tržaška 25
1000 Ljubljana
Apply as
Individual
Organisation
Student
Status: Not exceeded
Name: *
Email: *
Tip
Naziv pravne osebe
Ulica in hišna št.
Poštna št.
Kraj
Zavezanec za DDV
ID za DDV
Fakulteta
Telefon
Strinjam se s pogoji uporabe podatkov. *
  * Required field

Subscribe to newsletter