發(fā)布時(shí)間：2018-05-16 15:02

  本文選題：近紅外光譜 + 最小角回歸��； 參考：《光譜學(xué)與光譜分析》2017年06期

【摘要】：近紅外(NIR)光譜一般具有較多的波長(zhǎng)變量數(shù),對(duì)其直接或間接地進(jìn)行變量選擇是提高模型穩(wěn)定性能及預(yù)測(cè)性能的關(guān)鍵。最小角回歸(LAR)是一種相對(duì)較新和有效的機(jī)器學(xué)習(xí)算法,常用于進(jìn)行回歸分析和變量選擇。面向光譜建模應(yīng)用,提出一種LAR結(jié)合遺傳偏最小二乘法(GA-PLS)的變量選擇方法,可有效篩選出少數(shù)特征波長(zhǎng)點(diǎn)。首先在全光譜區(qū)利用LAR消除變量間的共線性得到初篩波長(zhǎng)點(diǎn),然后用GA-PLS對(duì)LAR篩選出的波長(zhǎng)點(diǎn)進(jìn)一步優(yōu)選從而得到最終建模用的特征波長(zhǎng)點(diǎn)。為驗(yàn)證本文方法的有效性,以藥片和汽油的近紅外光譜回歸分析作為應(yīng)用案例,對(duì)原光譜進(jìn)行預(yù)處理后,采用該方法進(jìn)行變量篩選,然后分別建模其中的活性成分含量和C10含量。結(jié)果顯示,在這兩個(gè)應(yīng)用中,最終優(yōu)化得到的特征波長(zhǎng)點(diǎn)數(shù)均只需七個(gè),而兩者的預(yù)測(cè)決定系數(shù)R2p分別達(dá)到0.933 9和0.951 9,與全光譜、無(wú)信息變量消除法(UVE)和連續(xù)投影算法(SPA)等方法相比,特征波長(zhǎng)點(diǎn)更少,同時(shí)R2p和預(yù)測(cè)均方根誤差RMSEP值更優(yōu)。因此,LAR結(jié)合GA-PLS,能有效地從近紅外光譜中選擇出信息變量從而減少建模波數(shù),提高預(yù)測(cè)精度,擁有較好的模型解釋性。該方法可為特定領(lǐng)域的專用光譜儀設(shè)計(jì)提供有效的波長(zhǎng)篩選工具。
[Abstract]:NIR spectra generally have a large number of wavelength variables. The selection of NIR spectra directly or indirectly is the key to improve the stability and prediction performance of the model. Minimum angle regression algorithm is a relatively new and effective machine learning algorithm, which is often used for regression analysis and variable selection. A variable selection method based on LAR combined with genetic partial least square method (GA-PLS) is proposed for spectral modeling, which can effectively screen a few characteristic wavelength points. In the whole spectrum region, the initial wavelength points are obtained by using LAR to eliminate the collinearity between variables, and then the characteristic wavelength points for the final modeling are obtained by the further optimization of the wavelength points selected by LAR by using GA-PLS. In order to verify the effectiveness of this method, the near infrared spectral regression analysis of tablets and gasoline was used as an application case. After pretreatment of the original spectrum, the method was used to screen the variables. Then, the content of active components and the content of C _ (10) were modeled respectively. The results show that, in these two applications, the number of characteristic wavelength points obtained by the final optimization is only seven, and the predictive determination coefficients R2p of the two methods are 0.933 9 and 0.951 9, respectively, which are in agreement with the full spectrum. Compared with the continuous projection algorithm (spa) and without information variable elimination (UVEE), there are fewer characteristic wavelength points, and R2P and the RMSEP value of RMS error are better. Therefore, Lar combined with GA-PLS can effectively select information variables from NIR spectra, thus reducing modeling wavenumber, improving prediction accuracy and having better model interpretation. This method can provide an effective wavelength screening tool for the design of special spectrometers in specific fields.
【作者單位】： 桂林電子科技大學(xué)電子工程與自動(dòng)化學(xué)院;北京郵電大學(xué)自動(dòng)化學(xué)院;
【基金】：國(guó)家自然科學(xué)基金項(xiàng)目(21365008,61562013) 廣西壯族自治區(qū)自然科學(xué)基金項(xiàng)目(2013GXNSFBA019279) 桂林電子科技大學(xué)研究生創(chuàng)新項(xiàng)目(GDYCSZ201474,GDYCSZ201478)資助
【分類號(hào)】：O657.33;R737.31

【相似文獻(xiàn)】

相關(guān)期刊論文 前2條

1 吳迪;汪志平;何勇;周子立;;iPLS-SPA變量選擇方法在螺旋藻粉無(wú)損檢測(cè)中的應(yīng)用[J];農(nóng)業(yè)工程學(xué)報(bào);2009年S2期

2 ;[J];;年期

相關(guān)會(huì)議論文 前3條

1 李洪東;梁逸曾;;高維數(shù)據(jù)變量選擇新方法研究[A];中國(guó)化學(xué)會(huì)第27屆學(xué)術(shù)年會(huì)第15分會(huì)場(chǎng)摘要集[C];2010年

2 徐登;范偉;梁逸曾;;紫外光譜結(jié)合變量選擇和偏最小二乘回歸同時(shí)測(cè)定水中重金屬鎘、鋅、鈷[A];中國(guó)化學(xué)會(huì)第29屆學(xué)術(shù)年會(huì)摘要集——第19分會(huì)：化學(xué)信息學(xué)與化學(xué)計(jì)量學(xué)[C];2014年

3 梁逸曾;李洪東;許青松;曹東升;張志敏;;灰色化學(xué)建模與模型集群分析——兼論過(guò)擬合、穩(wěn)健估計(jì)、變量選擇與模型評(píng)價(jià)[A];中國(guó)化學(xué)會(huì)第27屆學(xué)術(shù)年會(huì)第15分會(huì)場(chǎng)摘要集[C];2010年

相關(guān)博士學(xué)位論文 前2條

1 唐凱臨;變量選擇和變換的新方法研究[D];同濟(jì)大學(xué);2008年

2 唐麗娟;定量結(jié)構(gòu)活性相關(guān)性研究與高維微陣列數(shù)據(jù)分析中的化學(xué)計(jì)量學(xué)新算法[D];湖南大學(xué);2009年

相關(guān)碩士學(xué)位論文 前1條

1 龍旭霞;基于互信息的變量選擇方法研究[D];中南大學(xué);2013年

，

本文編號(hào)：1897335