Rape honey authenticity recognition method based on principal component analysis and gradient boosting decision tree

Abstract

The invention discloses a rape honey authenticity recognition method based on principal component analysis and a gradient boosting decision tree. Comprising the steps of collecting and preparing true and false honey samples and generating honey data, performing true and false labeling on honey data records to obtain a rape honey data set, performing dimensionality reduction on the rape honey data set by using a principal component analysis (PCA), constructing a rape honey true and false recognition model by using a gradient boosting decision tree (GBDT) algorithm, and evaluating the effectiveness of the model. And carrying out authenticity recognition on to-be-detected honey by using the rape honey authenticity recognition model. According to the method, the root-mean-square error and the AUC value are effectively improved, the data feature dimension, the model training time and the model complexity are reduced, the authenticity of the rape honey can be effectively and accurately recognized, the error caused by manually checking the spectrogram for authenticity recognition is avoided, and the accuracy of the authenticity recognition of the rape honey is improved.

Description

technical field [0001] The present invention relates to honey authenticity identification technology, in particular to a method for authenticity identification of rapeseed honey based on principal component analysis and gradient lifting decision tree Background technique [0002] Honey is a natural sweet substance that bees collect nectar from the flowers of flowering plants and fully brew in the hive. It has a strong smell and a pure and sweet taste. Honey is a sugar-based natural food, with glucose and fructose as its main components, which can be directly absorbed by the human body without enzymatic decomposition. It is also one of the most commonly used tonics and is deeply loved by consumers. [0003] However, the composition of honey is complex, and the composition content is related to the type of nectar source plant, flowering period, climate and storage and processing technology. At present, the national food safety standard GB / T18932.1-2002 and GB / T 18932.2-2002, ...

AI Technical Summary

Problems solved by technology

[0003] However, the composition of honey is complex, and the composition content is related to the type of nectar source plant, flowering period, climate, storage and processing technology, etc. At present, the domestic food safety national standard GB / T18932.1-2002 and GB / T 18932.2-2002, Using targeted detection techniques, such as mass spectrometry or chromatography, to determine the characteristic components of honey blends

However, counterfeiters can avoid the detection items in the standard in an organized way, so the current national standard can no longer meet the detection of honey adulteration or mixing at the present stage, and there is an urgent need to develop and establish a machine learning-based honey authenticity identification Useful ways

Images