Energy landscape paving with local search for global optimization of the BLN off-lattice model

2014 ◽  
Vol 64 (4) ◽  
pp. 603-610 ◽  
Author(s):  
Jingfa Liu ◽  
Weibo Huang ◽  
Wenjie Liu ◽  
Beibei Song ◽  
Yuanyuan Sun ◽  
...  
Keyword(s):  
Global Optimization ◽  
Local Search ◽  
Lattice Model ◽  
Energy Landscape ◽  
Energy Landscape Paving

Energy-landscape paving for prediction of face-centered-cubic hydrophobic-hydrophilic lattice model proteins

2013 ◽  
Vol 88 (5) ◽  
Author(s):  
Jingfa Liu ◽  
Beibei Song ◽  
Zhaoxia Liu ◽  
Weibo Huang ◽  
Yuanyuan Sun ◽  
...  
Keyword(s):  
Lattice Model ◽  
Energy Landscape ◽  
Face Centered Cubic ◽  
Face Centered ◽  
Energy Landscape Paving

scholarly journals Focused Crawler Strategy Based on Improved Energy Landscape Paving Algorithm

2020 ◽  
Author(s):  
Jingfa Liu ◽  
Wei Zhang ◽  
Zhihe Yang ◽  
Ziang Liu
Keyword(s):  
Global Optimization ◽  
Semantic Analysis ◽  
Energy Landscape ◽  
Optimization Method ◽  
Web Pages ◽  
Web Content ◽  
Local Optima ◽  
Link Structure ◽  
Focused Crawlers ◽  
Energy Landscape Paving

The traditional crawlers have difficulty in implementing semantic analysis. Therefore, the focused crawler technologies with topic preference characteristics have received many attentions in the recent years. To increase the precision of focused crawlers and prevent “topic drifting”, this paper adopts the comprehensive relevancy evaluation (CRE) of hyperlinks based on the combination of web content and link structure. In addition, the improved version of the energy landscape paving (ELP) algorithm that is a class of metropolis-sampling-based global optimization method is proposed to avoid the focused crawler falling into local optima. By incorporating the CRE strategy into the improved ELP, a novel focused crawler strategy denoted by IELP is proposed. The experimental results on rainstorm disasters domain show that the precision of the proposed focused crawler is obviously promoted compared to other focused crawlers in literature, illustrating the ability of the IELP to retrieve topic-related web pages.

STUDIES OF FINDING LOW ENERGY CONFIGURATIONS IN OFF-LATTICE PROTEIN MODELS

2006 ◽  
Vol 05 (03) ◽  
pp. 587-594 ◽  
Author(s):  
JINGFA LIU ◽  
WENQI HUANG
Keyword(s):  
Lower Energy ◽  
Gradient Descent ◽  
Energy Landscape ◽  
Three Dimensional ◽  
Ground States ◽  
Low Energy ◽  
Protein Models ◽  
Energy Configurations ◽  
Energy Landscape Paving ◽  
Lattice Protein Models

We studied two three-dimensional off-lattice protein models with two species of monomers, hydrophobic and hydrophilic. Low energy configurations in both models were optimized using the energy landscape paving (ELP) method and subsequent gradient descent. The numerical results show that the proposed methods are very promising for finding the ground states of proteins. For all sequences with lengths 13 ≤ n ≤ 55, the algorithm finds states with lower energy than previously proposed putative ground states.

Sign in / Sign up

Export Citation Format

Share Document