透過您的圖書館登入
IP:216.73.216.100
  • 學位論文

以大數據分析豬隻育種選拔

Big data analytics for the selective breeding in Pig

指導教授 : 黃正魁
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


種豬改良策略共有三種簡單而著名的策略,可以改善豬的遺傳組成進而影響改良方向,此三種策略如下: 1.利用雜交優勢。 2.導入來自不同種豬的新基因。 3.對特定性狀進行場內正向選拔。 在一個種豬群中,其任何特定性狀的表現都有變異,通常這種變異都屬常態分佈的,因此大部分動物都處於平均水平,而只有一少部分是特殊的。一般的育種改良認為選擇性能優秀的個體做為父母代,這樣得到的下一代之整體平均水平就會朝向正向改變,因此就必須透過選留的方式來正向的提升場內豬隻整體生產性能。本研究利用了過去收集的大量數據,以資料探勘工具-決策樹,來進行優良豬群特徵分析,提供未來育種改良的參考。

並列摘要


There are three uncomplicated but famous approaches to address the issue of sow breeding improvement. They can improve the genetic formation, leading to an impact on the quality of swine breeding. The three approaches are as follows. (1)Utilization of hybridization advantage (2)Introduction of novel genes from different sows (3)Positive selection from the inner field with respect to the specific traits In the third approach, there are of variances in the presentation of the specific traits. In general, these variances are belonging to the normal distribution. Therefore, most of swine are the above level for the specific traits; however, some are part of the outliers. Practictioners also choose the better parents to be as a pool for the breeding improvement. Accordingly, the next generation will tend to the positive way and improve the quality of swine breeding. As the argument above, there is a research problem we want to address; that is, how to determine the better parents for the next generation. In this study, we employ the big data from the past historical records of swine breeding and use the technique of data mining – decision tree to analyze the feature of the better parents. The analysis result can proffer the reference of breeding improvement.

參考文獻


呂明澤(2007)。運用資料挖礦技術進行影響良率學習之因素分析-以某半導體廠製程為例,國立清華大學工業工程與工程管理學系,碩士論文。
連德蓉(2013)。以決策樹機制分析客戶財務指標避免被倒帳風險之初,嶺東科技大學高階主管企業管理研究所,碩士論文。
張瀚文(2014)。應用決策樹於水庫防洪減淤最佳操作規則之研究,國立交通大學土木工程研究所,碩士論文。
張峻瑋(2014)。肉豬屠體分級與部位肉資訊系統,國立中興大學動物科學系,碩士論文。
簡禎富、李培瑞、彭誠湧(2003)。半導體製程資料特徵萃取與資料挖礦之研究.[Semiconductor Manufacturing Data Mining for Clustering and Feature Extraction],資訊管理學報,中華民國資訊管理學會。

延伸閱讀