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

切換: 標籤 | MARC模式 | ISBD

Validating a neural network-based on...

Liu, Yan.

FindBook

Google Book

Amazon

博客來

Validating a neural network-based online adaptive system.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Validating a neural network-based online adaptive system./
作者:	Liu, Yan.
面頁冊數:	98 p.
附註:	Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2159.
Contained By:	Dissertation Abstracts International66-04B.
標題:	Computer Science. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3170934
ISBN:	0542075660

Validating a neural network-based online adaptive system.
Liu, Yan.

Validating a neural network-based online adaptive system. - 98 p.

Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2159.

Thesis (Ph.D.)--West Virginia University, 2005.

Neural networks are popular models used for online adaptation to accommodate system faults and recuperate against environmental changes in real-time automation and control applications. However, the adaptivity limits the applicability of conventional verification and validation (V&V) techniques to such systems. We investigated the V&V of neural network-based online adaptive systems and developed a novel validation approach consisting of two important methods. (1) An independent novelty detector at the system input layer detects failure conditions and tracks abnormal events/data that may cause unstable learning behavior. (2) At the system output layer, we perform a validity check on the network predictions to validate its accommodation performance.

ISBN: 0542075660Subjects--Topical Terms:

626642
Computer Science.

Validating a neural network-based online adaptive system.
LDR:03006nmm 2200301 4500 001 1814707
005 20060719075906.5
008 130610s2005 eng d
020 $a 0542075660
035 $a (UnM)AAI3170934
035 $a AAI3170934
040 $a UnM $c UnM
100 1 $a Liu, Yan. $3 1067479
245 1 0 $a Validating a neural network-based online adaptive system.
300 $a 98 p.
500 $a Source: Dissertation Abstracts International, Volume: 66-04, Section: B, page: 2159.
500 $a Chair: Bojan Cukic.
502 $a Thesis (Ph.D.)--West Virginia University, 2005.
520 $a Neural networks are popular models used for online adaptation to accommodate system faults and recuperate against environmental changes in real-time automation and control applications. However, the adaptivity limits the applicability of conventional verification and validation (V&V) techniques to such systems. We investigated the V&V of neural network-based online adaptive systems and developed a novel validation approach consisting of two important methods. (1) An independent novelty detector at the system input layer detects failure conditions and tracks abnormal events/data that may cause unstable learning behavior. (2) At the system output layer, we perform a validity check on the network predictions to validate its accommodation performance.
520 $a Our research focuses on the Intelligent Flight Control System (IFCS) for NASA F-15 aircraft as an example of online adaptive control application. We utilized Support Vector Data Description (SVDD), a one-class classifier to examine the data entering the adaptive component and detect potential failures. We developed a "decompose and combine" strategy to drastically reduce its computational cost, from O(n 3) down to O( n32 log n) such that the novelty detector becomes feasible in real-time.
520 $a We define a confidence measure, the validity index, to validate the predictions of the Dynamic Cell Structure (DCS) network in IFCS. The statistical information is collected during adaptation. The validity index is computed to reflect the trustworthiness associated with each neural network output. The computation of validity index in DCS is straightforward and efficient.
520 $a Through experimentation with IFCS, we demonstrate that: (1) the SVDD tool detects system failures accurately and provides validation inferences in a real-time manner; (2) the validity index effectively indicates poor fitting within regions characterized by sparse data and/or inadequate learning. The developed methods can be integrated with available online monitoring tools and further generalized to complete a promising validation framework for neural network based online adaptive systems.
590 $a School code: 0256.
650 4 $a Computer Science. $3 626642
690 $a 0984
710 2 0 $a West Virginia University. $3 1017532
773 0 $t Dissertation Abstracts International $g 66-04B.
790 1 0 $a Cukic, Bojan, $e advisor
790 $a 0256
791 $a Ph.D.
792 $a 2005
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3170934