東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Evaluation of a new modeling tool to...

Shiels, Kerry.

Linked to FindBook

Google Book

Amazon

博客來

Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics./
Author:	Shiels, Kerry.
Description:	94 p.
Notes:	Source: Masters Abstracts International, Volume: 48-02, page: 0865.
Contained By:	Masters Abstracts International48-02.
Subject:	Agriculture, Horticulture. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1472637
ISBN:	9781109510812

Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics.
Shiels, Kerry.

Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics. - 94 p.

Source: Masters Abstracts International, Volume: 48-02, page: 0865.

Thesis (M.S.)--University of California, Davis, 2009.

There are a multitude of vineyard practices that can be manipulated in order to affect changes in the final chemical and sensory characteristics of a wine. Given the complex relationship between viticultural practices and wine characteristics, it is often difficult to predict the outcome and optimize growing conditions for key target attributes in wine (e.g. specific aromas, flavors or phenolic profiles).

ISBN: 9781109510812Subjects--Topical Terms:

1017832
Agriculture, Horticulture.

Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics.
LDR:03128nam 2200289 4500 001 1398150
005 20110907152349.5
008 130515s2009 ||||||||||||||||| ||eng d
020 $a 9781109510812
035 $a (UMI)AAI1472637
035 $a AAI1472637
040 $a UMI $c UMI
100 1 $a Shiels, Kerry. $3 1677024
245 1 0 $a Evaluation of a new modeling tool to aid in viticultural decision making based on desired wine characteristics.
300 $a 94 p.
500 $a Source: Masters Abstracts International, Volume: 48-02, page: 0865.
502 $a Thesis (M.S.)--University of California, Davis, 2009.
520 $a There are a multitude of vineyard practices that can be manipulated in order to affect changes in the final chemical and sensory characteristics of a wine. Given the complex relationship between viticultural practices and wine characteristics, it is often difficult to predict the outcome and optimize growing conditions for key target attributes in wine (e.g. specific aromas, flavors or phenolic profiles).
520 $a Neural networks can be used as a tool for evaluating complex nonlinear relationships among a large number of input variables. For this reason, neural network methodologies have been employed to model viticultural effects on the chemical profile of wines. In any viticulture trial, a long-term validation process is needed in order to evaluate the efficacy of any technique. Data collected from the 2000-2002 vintages in experimental vineyards in Oakville were used in this study to identify combinations of viticultural practices that have the most impact on the chemical characteristics of these Cabernet Sauvignon wines. The identified combinations of practices, along with data from the associated wines, were used to train neural network models.
520 $a In 2007 and 2008, Cabernet Sauvignon grapes were grown with five different viticultural treatments. These were subsequently made into wine in order to examine how well our models predicted the outcome of new combinations of viticultural practices. Treatments were purposely chosen to represent a wide range of predicted chemical profiles for these wines.
520 $a General regression neural networks predicted wine attributes well for two successive vintages. The more historical data used to train the networks, the better the predictions. In addition, effects of treatments, and interactions of multiple treatments were investigated and visualized graphically. Much of the variation not predicted by the model was attributed to variation in winemaking replications. When winemaking reps were combined, the model predicted the attribute well. This demonstrates the challenge of predicting wine qualities from viticulture inputs. However, using neural networks can provide a valuable tool for evaluating multiple inputs and predicting trends. This approach can be a useful tool in making viticulture decisions.
590 $a School code: 0029.
650 4 $a Agriculture, Horticulture. $3 1017832
650 4 $a Agriculture, Plant Culture. $3 1018669
690 $a 0471
690 $a 0479
710 2 $a University of California, Davis. $3 1018682
773 0 $t Masters Abstracts International $g 48-02.
790 $a 0029
791 $a M.S.
792 $a 2009
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1472637