Functional Nonparametric Predictions in Food Industry Using Near-Infrared Spectroscopy Measurement  

作　　者：Ibrahim M.Almanjahie Omar Fetitah Mohammed Kadi Attouch Tawfik Benchikh 

机构地区：[1]Department of Mathematics,College of Science,King Khalid University,Abha,62529,Saudi Arabia [2]Statistical Research and Studies Support Unit,King Khalid University,Abha,62529,Saudi Arabia [3]Laboratory of Statistics and Stochastic Processes University of Djillali Liabes BP 89,Sidi Bel Abbes,22000,Algeria [4]Medical Faculty,Djillali Liabes University BP 89,Sidi Bel Abbes,22000,Algeria

出　　处：《Computers, Materials & Continua》2023年第3期6307-6319,共13页计算机、材料和连续体（英文）

基　　金：This work is funded by the Deanship of Scientific Research at King Khalid University for funding this work through Large Groups Project under grant number RGP.2/132/43.

摘　　要：Functional statistics is a new technique for dealing with data thatcan be viewed as curves or images. Parallel to this approach, the Near-InfraredReflectance (NIR) spectroscopymethodology has been used in modern chemistryas a rapid, low-cost, and exact means of assessing an object’s chemicalproperties. In this research, we investigate the quality of corn and cookiedough by analyzing the spectroscopic technique using certain cutting-edgestatistical models. By analyzing spectral data and applying functional modelsto it, we could predict the chemical components of corn and cookie dough.Kernel Functional Classical Estimation (KFCE), Kernel Functional QuantileEstimation (KFQE), Kernel Functional Expectile Estimation (KFEE),Semi-Partial Linear Functional Classical Estimation (SPLFCE), Semi-PartialLinear Functional Quantile Estimation (SPLFQE), and Semi-Partial LinearFunctional Expectile Estimation (SPLFEE) are models used to accuratelyestimate the different quantities present in Corn and Cookie dough. Theselection of these functional models is based on their ability to constructa forecast region with a high level of confidence. We demonstrate that theconsidered models outperform traditional models such as the partial leastsquaresregression and the principal component regression in terms of predictionaccuracy. Furthermore, we show that the proposed models are morerobust than competing models such as SPLFQE and SPLFEE in the sensethat data heterogeneity has no effect on their efficiency.

关 键 词：Functional statistics NIR chemical component kernel estimation 

分 类 号：O657.33[理学—分析化学] TS201[理学—化学]