東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁 到查詢結果 [ null ]

切換: 標籤 | MARC模式 | ISBD

Pattern Mining and Anomaly Detection...

Karaca, Meserret,

FindBook

Google Book

Amazon

博客來

Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data // Meserret Karaca.
作者:	Karaca, Meserret,
面頁冊數:	1 electronic resource (73 pages)
附註:	Source: Dissertations Abstracts International, Volume: 83-12, Section: B.
Contained By:	Dissertations Abstracts International83-12B.
標題:	Industrial engineering. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29065268
ISBN:	9798802700624

Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data /
Karaca, Meserret,

Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data /Meserret Karaca. - 1 electronic resource (73 pages)

Source: Dissertations Abstracts International, Volume: 83-12, Section: B.

Discrete sequential data is the collection of an ordered series of discrete events or abstractions from a set of data points collected at different time periods. These types of processes are one of the most common and important process types encountered around us, in a variety of domains such as bio-informatics, healthcare management, intrusion detection, and banking processes. It is becoming customary to discover similarities and to detect the indicators for anomalies in a single variable and multivariate forms in both supervised and unsupervised settings. In this dissertation, we first use mainstream sequential frequent pattern mining methods to find similarities for 1) knowledge discovery and pattern extraction in multivariate data. We demonstrate the efficiency of this algorithm on standard datasets and then apply the method to a real healthcare problem. Next, we use graph transformations 2) to monitor and 3) to classify sequential network representations in order to detect anomalies in order to intervene and prevent failures for 2) single variable and 3) multivariate temporal sequential data. For single variable, the approach uses sequential graph transformation and monitors distance measures between consecutive subsequences given a window and a stride size for stationary and non-stationary data streams. We apply this method to synthetic and real-life data and compare it with current anomaly detection methods. In a multivariate setting, a classifier trains a series of feature vectors obtained with sequential graph transformation for multivariate temporal sequences in order to predict anomalous and normal sequences. We conduct a comparison analysis considering current time series classification algorithms as a benchmark. The analysis sheds light on how to deal with multivariate time intervals in deep learning algorithms. In conclusion, this dissertation aims to improve computational efficiency for memory and speed, and objective-oriented accuracy.

English

ISBN: 9798802700624Subjects--Topical Terms:

526216
Industrial engineering.
Subjects--Index Terms:

Discrete sequential data

Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data /
LDR:03431nmm a22004093i 4500 001 2437622
005 20251124054947.5
006 m o d
007 cr|nu||||||||
008 261011s2022 miu||||||m |||||||eng d
020 $a 9798802700624
035 $a (MiAaPQD)AAI29065268
035 $a AAI29065268
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Karaca, Meserret, $e author. $3 3828673
245 1 0 $a Pattern Mining and Anomaly Detection on Multivariate Discrete Sequential Data / $c Meserret Karaca.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2022
300 $a 1 electronic resource (73 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 83-12, Section: B.
500 $a Advisors: Alvarado, Michelle; Pardalos, Panos Committee members: Reisi-Gahrooei, Mostafa; Thai, My.
502 $b Ph.D. $c University of Florida $d 2022.
520 $a Discrete sequential data is the collection of an ordered series of discrete events or abstractions from a set of data points collected at different time periods. These types of processes are one of the most common and important process types encountered around us, in a variety of domains such as bio-informatics, healthcare management, intrusion detection, and banking processes. It is becoming customary to discover similarities and to detect the indicators for anomalies in a single variable and multivariate forms in both supervised and unsupervised settings. In this dissertation, we first use mainstream sequential frequent pattern mining methods to find similarities for 1) knowledge discovery and pattern extraction in multivariate data. We demonstrate the efficiency of this algorithm on standard datasets and then apply the method to a real healthcare problem. Next, we use graph transformations 2) to monitor and 3) to classify sequential network representations in order to detect anomalies in order to intervene and prevent failures for 2) single variable and 3) multivariate temporal sequential data. For single variable, the approach uses sequential graph transformation and monitors distance measures between consecutive subsequences given a window and a stride size for stationary and non-stationary data streams. We apply this method to synthetic and real-life data and compare it with current anomaly detection methods. In a multivariate setting, a classifier trains a series of feature vectors obtained with sequential graph transformation for multivariate temporal sequences in order to predict anomalous and normal sequences. We conduct a comparison analysis considering current time series classification algorithms as a benchmark. The analysis sheds light on how to deal with multivariate time intervals in deep learning algorithms. In conclusion, this dissertation aims to improve computational efficiency for memory and speed, and objective-oriented accuracy.
546 $a English
590 $a School code: 0070
650 4 $a Industrial engineering. $3 526216
650 4 $a Systems science. $3 3168411
653 $a Discrete sequential data
653 $a Pattern extraction
653 $a Sequential network representations
690 $a 0546
690 $a 0790
710 2 $a University of Florida. $b Industrial and Systems Engineering. $e degree granting institution. $3 3828674
720 1 $a Alvarado, Michelle $e degree supervisor.
720 1 $a Pardalos, Panos $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 83-12B.
790 $a 0070
791 $a Ph.D.
792 $a 2022
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=29065268