Transgenic soybean classification and identification method based on PCA-SVM algorithm

Abstract

The invention discloses a transgenic soybean classification and identification method based on a PCA-SVM algorithm. Spectroscopic data obtained through absorption of transgenic soybean biomacromolecules to terahertz time-domain spectroscopy is utilized, principal component analysis is firstly carried out, principal component factors with the number of a single digit are obtained after dimension reduction processing, multiple samples are classified and identified through a support vector machine algorithm, and finally a classifying result figure of the soybean samples is obtained. The result also proves that the classifying result accuracy rate of the method on transgenic soybeans is 100%. The dimension reduced spectroscopic data and the support vector machine algorithm are utilized, a large number of soybean samples can be processed at the same time, and the samples can be classified and identified fast, efficiently and accurately.

Description

technical field [0001] The invention relates to a method for identifying transgenic soybeans, in particular to a method for classifying and identifying transgenic soybeans based on a THz-TDS PCA-SVM model. Background technique [0002] The toxicity and potential safety hazards of genetically modified soybeans have always been controversial. Edible soybean oil has undergone complex processing links, and DNA degradation and damage are extremely serious, which has brought great difficulties to genetically modified soybeans, and false negatives are prone to occur in detection. Therefore, before the extraction begins, using the property of DNA soluble in aqueous solution, add a certain volume of TE solution to the edible oil for washing. This step is crucial for the successful extraction of DNA. By improving DNA extraction methods, designing quality control primers for endogenous genes, controlling the length of amplified fragments, and using different detection techniques, false...

AI Technical Summary

Problems solved by technology

[0002] The toxicity and potential safety hazards of genetically modified soybeans have always been controversial. Edible soybean oil undergoes complex processing links, and DNA degradation and damage are extremely serious, which brings great difficulties to the detection of genetically modified soybeans, and false negatives are prone to occur in the detection

[0005] Classification and identification of transgenic soybeans using THz-TDS-based support vector machine (SVM) algorithm is still a blank research field

Images