当前位置: X-MOL 学术Int. J. Inf. Technol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Estimation of crop production using machine learning techniques: a case study of J&K
International Journal of Information Technology Pub Date : 2021-05-05 , DOI: 10.1007/s41870-021-00653-7
Jyoti Mahajan , Kriti Banal , Samridhi Mahajan

A large part of the world population has been engaged in agriculture and its related activities since ancient times. Farmers have usually relied on their own experiences to make any decision about agriculture or farm management. As agriculture is greatly affected by unpredictable weather conditions, it becomes difficult to take decisions with total confidence, and many a times farmers have to suffer great losses. If we could predict with certain accuracy, the future yield of crops, it would greatly help the farmers in deciding which crops to focus on, to obtain maximum profits. This paper presents a crop yield prediction system using machine learning algorithms (decision trees, random forest, support vector regressor, gradient boosting) which can be used to predict the yields of major crops produced in regions of Jammu and Kashmir in India. Historical production and meteorological data was collected and processed for analysis and applying ML algorithms. The models were validated by testing on data they have not trained on. Among the models used, decision tree regressor predicts crop production with the greatest accuracy of 96% while SVR performs the poorest with an accuracy of 89%.



中文翻译:

使用机器学习技术估算农作物产量:J&K案例研究

自古以来,世界上有很大一部分人口从事农业及其相关活动。农民通常依靠自己的经验对农业或农场管理做出任何决定。由于农业受到不可预测的天气条件的极大影响,因此很难完全自信地做出决定,很多时候农民不得不遭受巨大损失。如果我们能够以一定的准确度预测未来的农作物产量,那将极大地帮助农民决定要重点种植哪种农作物,从而获得最大的利润。本文提出了一种使用机器学习算法(决策树,随机森林,支持向量回归,梯度增强)的农作物产量预测系统,该系统可用于预测印度查mu和克什米尔地区主要农作物的产量。收集并处理了历史生产和气象数据,以进行分析和应用ML算法。通过对尚未训练的数据进行测试来验证模型。在使用的模型中,决策树回归器以96%的最高准确度预测作物产量,而SVR以89%的准确度预测最差的作物产量。

更新日期:2021-05-05
down
wechat
bug