Optimizing prediction of human assessments of dairy odors using input variable selection
使用輸入變量選擇優(yōu)化人類乳品氣味評估的預(yù)測
Fangle Changa, Paul H. Heinemannb,?
a Department of Agricultural and Biological Engineering, The Pennsylvania State University, 105 Agricultural Engineering Building, University Park, PA 16802, USA
b Department of Agricultural and Biological Engineering, Pennsylvania State University, 105 Agricultural Engineering Building, University Park, PA 16802, USA
A B S T R A C T
Use of instruments instead of human panels to assess odors can make the collection and measurement process more efficient and reliable. Odor-emitting samples from dairy farms, including manure, feed, and bedding materials, were collected and assessed by an electronic nose and a human panel. Artificial neural networks based on the Levenberg-Marquardt Back-propagation algorithm were used to build prediction models to predict human response to odor pleasantness. Feature selection methods, including Forward Selection (FS), Gamma Test (GT), and Principal Component Analysis (PCA), were applied to reduce the dimensionality of the measurements, potentially eliminating noise. Out of the 28 variable candidates (eNose sensors), 10 variables were selected when PCA was applied, and 16 variables were selected when either FS or GT approaches were applied. The model developed using GT provided the lowest mean square error of 0.56 (2.5%) hedonic scale units for separate validation. The GT-based model was able to predict the human assessments within 10% of the target for 81% of the independent validation samples and within 5% of the target for 63% of the independent validation samples.
使用儀器而不是人體感官來評估氣味，可以使收集和測量過程更加有效和可靠。從奶牛場采集臭味樣本，包括糞便、飼料和床上用品，并通過電子鼻和人體感官進(jìn)行評估。采用基于Levenberg-Marquardt反向傳播算法的人工神經(jīng)網(wǎng)絡(luò)建立預(yù)測模型，預(yù)測人類對氣味愉悅的反應(yīng)。采用前向選擇(FS)、伽瑪檢驗(yàn)(GT)和主成分分析(PCA)等特征選擇方法來降低測量的維數(shù)，有可能消除噪聲。在28個(gè)變量候選(eNose傳感器)中，當(dāng)應(yīng)用PCA時(shí)選擇10個(gè)變量，當(dāng)應(yīng)用FS或GT方法時(shí)選擇16個(gè)變量。使用GT開發(fā)的模型為單獨(dú)驗(yàn)證提供了0.56(2.5%)標(biāo)度單位的小均方誤差?；贕T的模型能夠預(yù)測81%獨(dú)立驗(yàn)證樣本的目標(biāo)值的10%范圍內(nèi)的人類評估，63%獨(dú)立驗(yàn)證樣本的目標(biāo)值的5%范圍內(nèi)的人類評估。
會(huì)員1.png)
