BEGIN:VCALENDAR VERSION:2.0 PRODID:-//jEvents 2.0 for Joomla//EN CALSCALE:GREGORIAN METHOD:PUBLISH BEGIN:VEVENT UID:2f595f426cfb0d824c60a631ca93a46b CATEGORIES:Lectures & Presentations CREATED:20131016T142225 SUMMARY:Anomaly detection methods and applications LOCATION:SBA Research gGmbH - Wien DESCRIPTION:Speaker: Dalia Kriksciuniene, Assoc.prof.dr. (ERCIM researcher at Masaryk U niversity, Brno, Czech)\nTitle: Anomaly detection methods and applications\ nAbstract:\n \n \nNumerous statistical, econometrical and intelligent metho ds are researched in the scientific works and industrial applications for f inding repeating patterns and approximated distribution rules of data serie s. The main problem experienced by the scientific research is lack of regul arities and rather chaotic nature of real data. That hinders performance of the forecasting methods based on finding regularities in data distribution , limits their application to analytics of historical data repositories, an d reveals high unreliability of the models to react to occurring external i mpacts for real-time data streams.\n The current research aims to explore a pplication of computational analysis methods for detecting behaviour anomal ies. The methods and the principles of anomaly detection are analysed from the theoretical perspective and evaluated for their ability to serve for an alysis, forecasting and designing optimal strategies for the enterprises op eration in various application domains. They include analysis of anomalies of the financial markets, caused by changes of investor activeness due to different calendar effects, impacts of various media announcements and news , fluctuations of market situations, including crises and bubbles. The anom aly detection is important in the areas of facility management based on sen sor data system, detection of network intrusions and performance failures. \nThe results of currently published research cover several methods for det ecting behaviour anomalies including information efficiency evaluation meth ods (Shannon's entropy, Hurst exponent), event impacts explored by the meth ods of interrupted time series, and evaluation of binary clustering algorit hms. \nThe research results are evaluated by discussing the potential power of explored methods for detection behaviour anomalies in the expanding res earch area of Big Data. \n \n \n\nRead more... (http://www.sba-research.or g/) X-ALT-DESC;FMTTYPE=text/html:
Speaker: Da lia Kriksciuniene, Assoc.prof.dr. (ERCIM researcher at Masaryk University, Brno, Czech) Title: Anomaly detection methods and applications Abstra ct:
Numerous statistical , econometrical and intelligent methods are researched in the scientific wo rks and industrial applications for finding repeating patterns and approxim ated distribution rules of data series. The main problem experienced by the scientific research is lack of regularities and rather chaotic nature of r eal data. That hinders performance of the forecasting methods based on find ing regularities in data distribution, limits their application to analytic s of historical data repositories, and reveals high unreliability of the mo dels to react to occurring external impacts for real-time data streams. The current research aims to explore application of comput ational analysis methods for detecting behaviour anomalies. The methods and the principles of anomaly detection are analysed from the theoretical pers pective and evaluated for their ability to serve for analysis, forecasting and designing optimal strategies for the enterprises operation in various a pplication domains. They include analysis of anomalies of the financial markets, caused by changes of investor activeness due to differe nt calendar effects, impacts of various media announcements and news, fluct uations of market situations, including crises and bubbles. The anomaly det ection is important in the areas of facility management based on sensor dat a system, detection of network intrusions and performance failures. The results of currently published research cover several methods for detec ting behaviour anomalies including information efficiency evaluation method s (Shannon's entropy, Hurst exponent), event impacts explored by the method s of interrupted time series, and evaluation of binary clustering algorithm s. The research results are evaluated by discussing the potential po wer of explored methods for detection behaviour anomalies in the expanding research area of Big Data.
| ||
Read more... |
||