tartary buckwheat flour is esteemed for its high nutritional value and price, yet it is susceptible to adulteration in the market. Near-infrared spectroscopy (NiRs), commonly utilized for nutrient content detection, has recently been applied to authenticate food products this study collected near-infrared spectral data from adulterated samples of whole wheat flour, oat flour, soybean flour, barley flour, and sorghum flour in tartary buckwheat. Utilizing partial least squares regression (PLsR), support vector regression (sVR), and back-propagation neural network (BPNN), we predicted the adulterant content in tartary buckwheat. the results demonstrated that the BPNN algorithm, with an determination coefficient of prediction exceeding 0.97 and root mean square error of prediction below 0.02, surpassed the PLsR and sVR models in predicting adulterated crop flour, showcasing superior accuracy and generalization capabilities. this integration of NiRs and BPNN proved effective for the quantitative analysis of crop flours in tartary buckwheat, exhibiting robust predictive performance and rapid detection of adulteration in agricultural products.