基于支持向量机奶牛饲粮能量消化预测方法  

Predicting method for diet energy digestion of dairy cows using support vector machine

作　　者：付强[1] 沈维政[1,2] 魏晓莉 张永根[3] 辛杭书 苏中滨[1] 赵春江[1,4] FU Qiang;SHEN Wei-zheng;WEI Xiao-li

机构地区：[1]东北农业大学电气与信息学院,黑龙江哈尔滨150030 [2]农业农村部生猪养殖设施工程重点实验室,黑龙江哈尔滨150030 [3]东北农业大学动物科学技术学院,黑龙江哈尔滨150030 [4]国家农业信息化工程技术研究中心,北京100097

出　　处：《饲料研究》2019年第7期47-51,共5页Feed Research

基　　金：“十三五”国家重点研发计划子课题“奶牛、肉牛信息感知及环境智能控制系统的研制与验证(项目编号:2016YFD0700204-02)”;国家奶牛产业技术体系(项目编号:CARS-36);中国博士后科学基金(项目编号:2017M611346);黑龙江省自然科学基金(项目编号:C2018018)

摘　　要：文章以预测奶牛饲粮消化能(DE)和能量消化率(ED)为目标,提出一种基于支持向量机(SVM)的预测方法,SVM是一种典型的非参数机器学习模型,无需提前对预测模型做任何假设,仅通过对训练样本数据的学习,便能拟合出最接近实际的函数。与传统LR方法和其他非参数机器学习模型RBF-ANN的对比测试结果表明,基于SVM的预测模型在预测精度上优于其他方法,特别是在小样本条件下比传统的LR方法预测精度更高,可以作为对奶牛饲粮DE和ED进行预测的新型参考方法。In the study, a support vector machine(SVM) technique was proposed to predict the indicators of digestible energy(DE) and energy digestibility(ED) of dairy cows. SVM is a typical non-parametric machine learning model that does not require any specific assumptions about the regression function in advance and only by learning the training sample data, then it can fit the function closest to the actual in most cases. To evaluate prediction accuracy effectively, the SVM technique was compared with the traditional LR method and other non-parametric model such as radial basis function artificial neural network(RBF-ANN) method. The prediction results indicate that the proposed SVM-based prediction technique is superior to other methods in most metrics. In particular, it has higher prediction accuracy than traditional LR method under small samples conditions, and can be used as a new reference method for predicting DE and ED indicators of dairy cows.

关 键 词：奶牛饲粮 消化能预测 能量消化率预测 非参数模型 支持向量机 

分 类 号：S811[农业科学—畜牧学]