Analysis And Evaluation Of Tobacco Planting Soil Nutrients Of Longhui County Based On Machine Learning

Longhui County is one of the tobacco planting areas in Hunan Province.The soil nutrients in different tobacco planting areas have certain regional features.In order to improve the scientific management and intelligent decision-making of tobacco planting soil,this paper integrates state of art methods and technologies including machine learning,statistics,decision support systems and databases,and conducts an analysis and evaluation study in tobacco planting soil nutrients of Longhui County.The main research contents are as follows:(1)According to the regional features of Longhui County tobacco planting area,applying with international soil quality rating standards and Hunan tobacco planting soil nutrient content rating standards,we analyzed soil features,the overall features of soil nutrient content,and the distribution of nutrient suitability of 19 soil nutrients elements in 67 soil example data collected from tobacco planting area of Longhui County by using descriptive statistical methods.The experimental results showed that the tobacco planting areas in Longhui County were mainly distributed in hilly areas,with moderate soil erosion,and the soil was mainly loamy clay.The nutrient indicators such as PH,alkaline nitrogen,fast-acting potassium,total nitrogen,total potassium,organic matter,available manganese,available zinc,available molybdenum,and cation exchange amount is moderate and content ranking distribution is appropriate.The nutrient indicators such as total phosphorus,exchangeable magnesium,and chloride ions are relatively low,and the content ranking distribution is low or very low.The nutrient indicators such as available iron,available copper,available boron,available sulfur,and exchangeable calcium are extremely high or high.(2)Based on the Apriori association rule analysis machine learning algorithm,the tobacco planting areas in Longhui County was analyzed.We obtained prediction model of tobacco planting nutrient indicators,revealed soil nutrients relationships and differences.And we also integrated decision support system and database technology to design a knowledge base of soil nutrients for tobacco planting,which provided scientific decision support for the intelligent analysis of nutrients in tobacco planting soil.(3)Pearson correlation analysis was applied on the soil nutrient index data of tobacco planting areas in Longhui County.The experimental results showed that there was a correlation between the nutrient indicators of the tobacco planting areas,and some of them showed a high correlation.We also used the unsupervised machine learning principal component analysis(PCA)dimension reduction evaluation method to evaluate the soil nutrients of the tobacco planting areas in Longhui County comprehensively.The results showed that among the ten towns in Longhui County,Zhouwang Town had the highest comprehensive evaluation score and Hetian Town had the lowest score.The research work provides a valuable reference for the rational usage of soil and fertilizer management in tobacco planting areas.