Tobacco leaf quality grade classification prediction method based on principal component analysis and super learning

Abstract

The invention discloses a tobacco leaf quality grade classification prediction method based on principal component analysis and super learning. The method comprises the following steps: 1) grouping tobacco leaf quality data samples according to set index categories; (2) principal component analysis is conducted on index data in each index data set, dimension reduction is conducted on the data, and correlation is eliminated; 3) training each basic learning algorithm in the super learning framework by using each processed index data set to obtain a first-level classification prediction model; 4) selecting verification data and inputting the verification data into the corresponding first-level classification prediction model to obtain a classification prediction result; 5) taking each classification prediction result as input data of a meta-learner in a super learning framework to train the meta-learner, obtaining an optimized weight combination of each first-level classification prediction model, and creating a super learning model; and 6) inputting the index data of the to-be-recognized tobacco leaf quality data into the super learning model to obtain a tobacco leaf quality grade classification prediction result of the to-be-recognized tobacco leaf quality data.

Description

technical field [0001] The invention relates to a method for classifying and predicting tobacco leaf quality grades based on machine learning, in particular to a method for classifying and predicting tobacco leaf quality grades based on principal component analysis and super learning. Background technique [0002] Tobacco leaf is an important raw material for the tobacco industry. The relationship between the appearance, physical properties, chemical composition, sensory and other quality indicators of tobacco leaves has been the focus of many studies, and has a direct impact on the quality of cigarette products. my country is a big country in the cultivation, production and consumption of tobacco leaves. Due to the influence of climate, soil, regional environment and varieties, planting measures, planting parts and curing techniques, the quality of tobacco leaves is quite different. To grasp the quality dynamics of tobacco leaves and determine the quality grade of tobacco l...

AI Technical Summary

Problems solved by technology

[0008] Aiming at the problems that the evaluation of tobacco leaf quality grades involves many indicators, subjective and objective differences, etc., which lead to large differences in evaluation results, and the prediction accuracy of tobacco leaf quality grade classification is not high, the present invention provides a tobacco leaf quality grade classification based on principal component analysis and super learning Prediction method, which can effectively improve the accuracy of tobacco leaf quality grade classification prediction

Images