The covariance selection quality for graphs with junction trees through AUC bounds

2016 ◽  
Author(s):  
Navid Tafaghodi Khajavi ◽  
Anthony Kuh
Keyword(s):  
Covariance Selection ◽  
Junction Trees

scholarly journals Long-Lasting Insecticide Netting for Protecting Tree Stems from Attack by Ambrosia Beetles (Coleoptera: Curculionidae: Scolytinae)

Insects ◽  
2019 ◽  
Vol 11 (1) ◽  
pp. 8
Author(s):  
Christopher M. Ranger ◽  
Christopher T. Werle ◽  
Peter B. Schultz ◽  
Karla M. Addesso ◽  
Jason B. Oliver ◽  
...  
Keyword(s):  
Field Conditions ◽  
Ambrosia Beetle ◽  
Ambrosia Beetles ◽  
Cercis Canadensis ◽  
Fine Mesh ◽  
Junction Trees ◽  
Eastern Redbud ◽  
Wood Boring ◽  
Xylosandrus Germanus

Ambrosia beetles (Coleoptera: Curculionidae: Scolytinae) are destructive wood-boring insects of horticultural trees. We evaluated long-lasting insecticide netting for protecting stems against ambrosia beetles. Container-grown eastern redbud, Cercis canadensis, trees were flood-stressed to induce ambrosia beetle attacks, and deltamethrin-treated netting was wrapped from the base of the stem vertically to the branch junction. Trees were deployed under field conditions in Ohio, Virginia, Tennessee, and Mississippi with the following treatments: (1) flooded tree; (2) flooded tree with untreated netting; (3) flooded tree with treated ‘standard mesh’ netting of 24 holes/cm2; (4) flooded tree with treated ‘fine mesh’ netting of 28 holes/cm2; and/or (5) non-flooded tree. Treated netting reduced attacks compared to untreated netting and/or unprotected trees in Mississippi in 2017, Ohio and Tennessee in 2018, and Virginia in 2017–2018. Inconsistent effects occurred in Mississippi in 2018. Fewer Anisandrus maiche, Xylosandrus germanus, and Xyleborinus saxesenii were dissected from trees deployed in Ohio protected with treated netting compared to untreated netting; trees deployed in other locations were not dissected. These results indicate long-lasting insecticide netting can provide some protection of trees from ambrosia beetle attacks.

Abstract B2-35: Efficient exploration of multi-cancer networks by generalized covariance selection and interactive web content

2015 ◽  
Author(s):  
Teresia Kling ◽  
Patrik Johansson ◽  
Jose Sanchez ◽  
Voichita D. Marinescu ◽  
Rebecka Jornsten ◽  
...  
Keyword(s):  
Web Content ◽  
Covariance Selection ◽  
Generalized Covariance ◽  
Efficient Exploration ◽  
Cancer Networks

scholarly journals Multiblock Analysis Applied to TD-NMR of Butters and Related Products

2020 ◽  
Vol 10 (15) ◽  
pp. 5317
Author(s):  
Jean-Michel Roger ◽  
Silvia Mas Garcia ◽  
Mireille Cambert ◽  
Corinne Rondeau-Mouro
Keyword(s):  
Solid Phase ◽  
Fat Content ◽  
Prediction Errors ◽  
Chemometric Methods ◽  
Covariance Selection ◽  
T1 Relaxation ◽  
Repeatability Error ◽  
Measurement Protocol ◽  
Relaxation Measurements ◽  
Better Than

This work presents a novel and rapid approach to predict fat content in butter products based on nuclear magnetic resonance longitudinal (T1) relaxation measurements and multi-block chemometric methods. The potential of using simultaneously liquid (T1L) and solid phase (T1S) signals of fifty samples of margarine, butter and concentrated fat by Sequential and Orthogonalized Partial Least Squares (SO-PLS) and Sequential and Orthogonalized Selective Covariance Selection (SO-CovSel) methods was investigated. The two signals (T1L and T1S) were also used separately with PLS and CovSel regressions. The models were compared in term of prediction errors (RMSEP) and repeatability error (σrep). The results obtained from liquid phase (RMSEP ≈ 1.33% and σrep≈ 0.73%) are better than those obtained with solid phase (RMSEP ≈ 5.27% and σrep≈ 0.69%). Multiblock methodologies present better performance (RMSEP ≈ 1.00% and σrep≈ 0.47%) and illustrate their power in the quantitative analysis of butter products. Moreover, SO-Covsel results allow for proposing a measurement protocol based on a limited number of NMR acquisitions, which opens a new way to quantify fat content in butter products with reduced analysis times.

Sign in / Sign up

Export Citation Format

Share Document