Method for verifying feasibility of spectral data model transfer algorithm

Abstract

The invention discloses a method for verifying feasibility of a spectral data model transfer algorithm. The method comprises the following steps: classifying to-be-tested samples into a calibration set and a prediction set; carrying out host detection to obtain host correction set spectral data and host prediction set spectral data, and carrying out slave detection to obtain slave detection set spectral data; performing PLS algorithm modeling by using the spectral data of the host calibration set, and predicting the spectral data of the host prediction set by using the model to obtain a content prediction value M; processing the spectral data of the slave prediction set by adopting a model transfer algorithm to obtain brand-new spectral data of the slave prediction set, and carrying out predicting by using a model to obtain a content prediction value N; and calculating a difference value between the predicted value M and the predicted value N, and if an absolute value of the differencevalue is smaller than a threshold value, determining that the model transfer algorithm is feasible. The feasibility of the model transfer algorithm can be quickly verified, and an optimal model transfer algorithm is selected for different samples and different near-infrared spectrometers.

Description

technical field [0001] The invention relates to the technical field of spectral data model transfer, in particular to a method for verifying the feasibility of a spectral data model transfer algorithm. Background technique [0002] The spectrum contains a wealth of material information, and the spectral analysis technology has the advantages of non-destructive and fast, and has been widely used in agriculture, food, industry and other fields. Modeling with spectral data and related chemical values ​​can quickly realize the prediction of chemical values ​​of unknown samples, but this spectral model has certain limitations in predicting unknown samples, and can only predict unknown samples within a certain range. Different temperatures, different instruments, different measurement conditions, and spectra of samples in different regions will lead to inaccurate prediction results. To solve this problem, one method is to collect sample spectra and chemical values ​​to re-establi...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a method for verifying the feasibility of the spectral data model transfer algorithm, which is used to solve the problem of using the optimized model transfer algorithm itself to improve the prediction ability after model transfer in the prior art but does not consider the matching between the model transfer algorithm and the current model Problems that lead to poor predictions

Images