Construction method and application of highland barley selenium content prediction model based on binary linear regression

Abstract

The invention discloses a construction method and application of a highland barley selenium content prediction model based on binary linear regression. The technical scheme adopted is as follows: perform binary linear regression on the selenium content in highland barley grain samples with the soil total selenium content and soil available selenium content in soil samples to obtain the prediction model for highland barley grain selenium content. In the model, highland barley grain selenium content = 1.4420 × soil available selenium content - 0.00674; where the units of highland barley selenium content and soil available selenium content are both mg / kg. The invention proposes a model for predicting the corresponding highland barley grain selenium content based on the soil available selenium content in highland barley producing areas, with a goodness of fit of 0.4745, reaching an extremely significant level; the prediction model has good reliability and stability.

Images