Browsing by Author "David, Esther"

Now showing 1 - 2 of 2

A study on anomaly detection ensembles
(2017-05-01)
Chiang, Alvin
;
David, Esther
;
Lee, Yuh-Jye
;
Leshem, Guy
;
Yeh, Yi-Ren
An anomaly, or outlier, is an object exhibiting differences that suggest it belongs to an as-yet undefined class or category. Early detection of anomalies often proves of great importance because they may correspond to events such as fraud, spam, or device malfunctions. By automating the creation of a ranking or list of deviations, we can save time and decrease the cognitive overload of the individuals or groups responsible for responding to such events. Over the years many anomaly and outlier metrics have been developed. In this paper we propose a clustering-based score ensembling method for outlier detection. Using benchmark datasets we evaluate quantitatively the robustness and accuracy of different ensemble strategies. We find that ensembling strategies offer only limited value for increasing overall performance, but provide robustness by negating the influence of severely underperforming models. (C) 2017 Elsevier B.V. All rights reserved.
Sparse Sampling for Sensing Temporal Data - Building an Optimized Envelope
(2016-01-01)
Domb, Menachem
;
Leshem, Guy
;
Bonchek-Dokow, Elisheva
;
David, Esther
;
Lee, Yuh-Jye
IoT systems collect vast amounts of data which can be used in order to track and analyze the structure of future recorded data. However, due to limited computational power, bandwith, and storage capabilities, this data cannot be stored as is, but rather must be reduced in such a way so that the abilities to analyze future data, based on past data, will not be compromised. We propose a parameterized method of sampling the data in an optimal way. Our method has three parameters-an averaging method for constructing an average data cycle from past observations, an envelope method for defining an interval around the average data cycle, and an entropy method for comparing new data cycles to the constructed envelope. These parameters can be adjusted according to the nature of the data, in order to find the optimal representation for classifying new cycles as well as for identifying anomalies and predicting future cycle behavior. In this work we concentrate on finding the optimal envelope, given an averaging method and an entropy method. We demonstrate with a case study of meteorological data regarding El Ninio years.