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

From networks to proteins: Modeling ...

Chen, Yin.

FindBook

Google Book

Amazon

博客來

From networks to proteins: Modeling and optimization with Markovian models.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	From networks to proteins: Modeling and optimization with Markovian models./
作者:	Chen, Yin.
面頁冊數:	110 p.
附註:	Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 7060.
Contained By:	Dissertation Abstracts International71-11B.
標題:	Statistics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3430355
ISBN:	9781124295886

From networks to proteins: Modeling and optimization with Markovian models.
Chen, Yin.

From networks to proteins: Modeling and optimization with Markovian models. - 110 p.

Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 7060.

Thesis (Ph.D.)--Boston University, 2011.

Markov chain and Markov random field (MRF) theory provides a consistent way to model context-dependent entities and correlated features. It is used in conjunction with statistical decision and estimation theories to formulate objective functions in terms of established optimality principles. This dissertation applies MRF modeling and optimization in two areas: network security and protein docking.

ISBN: 9781124295886Subjects--Topical Terms:

517247
Statistics.

From networks to proteins: Modeling and optimization with Markovian models.
LDR:03451nam 2200313 4500 001 1396713
005 20110620110233.5
008 130515s2011 ||||||||||||||||| ||eng d
020 $a 9781124295886
035 $a (UMI)AAI3430355
035 $a AAI3430355
040 $a UMI $c UMI
100 1 $a Chen, Yin. $3 1675508
245 1 0 $a From networks to proteins: Modeling and optimization with Markovian models.
300 $a 110 p.
500 $a Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 7060.
500 $a Adviser: Ioannis Paschalidis.
502 $a Thesis (Ph.D.)--Boston University, 2011.
520 $a Markov chain and Markov random field (MRF) theory provides a consistent way to model context-dependent entities and correlated features. It is used in conjunction with statistical decision and estimation theories to formulate objective functions in terms of established optimality principles. This dissertation applies MRF modeling and optimization in two areas: network security and protein docking.
520 $a The first problem is network anomaly detection. The objective is to detect statistically significant temporal or spatial changes in either the underlying process being monitored or the network operation itself. These changes may point to faults, threats, misbehavior, or other anomalies that require intervention. A new statistical anomaly detection framework is introduced that uses Markov models to characterize the "normal" behavior of the network. A series of Markov models are developed, including a node-level model, two tree-indexed models, and a networked Markov chain model to capture temporal and spatial features of network evolution. Large deviations techniques are employed to compare the empirical distribution (estimated from past activity traces) with its most recent empirical measure. Optimal decision rules are developed for each model to identify anomalies in recent activities. Simulation results validate the effectiveness of the proposed anomaly detection algorithms.
520 $a The second application lies in bioinformatics where the objective is to computationally predict the structure of a complex of two interacting proteins. In the refinement stage of a multi-stage protein docking process, side-chains are extracted from the docking interface and positioned optimally by minimizing an energy function which models interactions among atoms. This combinatorial optimization problem is formulated as a maximum weighted independent set (MWIS) problem. The MWIS problem is NP-hard, yet, the thesis develops a distributed message passing algorithm that can produce effective feasible solutions. The approach is tested on a benchmark of 17 proteins and the results indicate that optimal side-chain positioning significantly improves the binding energy landscape, increasing the correlation between energy and root mean square deviation (RMSD) from the native structure. This new side-chain positioning procedure has the potential to accelerate refinement algorithms seeking very accurate predictions of the native structure and thus serve as an important step in computational docking protocols.
590 $a School code: 0017.
650 4 $a Statistics. $3 517247
650 4 $a Biology, Bioinformatics. $3 1018415
650 4 $a Operations Research. $3 626629
690 $a 0463
690 $a 0715
690 $a 0796
710 2 $a Boston University. $3 1017454
773 0 $t Dissertation Abstracts International $g 71-11B.
790 1 0 $a Paschalidis, Ioannis, $e advisor
790 $a 0017
791 $a Ph.D.
792 $a 2011
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3430355