Forecasting Tea Production in the Context of Bangladesh Utilizing Machine Learning

Abstract

One of the most popular drinks, second only to water, is tea. Bangladesh ranks as the world’s 10th-largest manufacturer of tea. Tea has a big impact on poverty reduction, rural development, and nutrition security. Over the past ten years, Bangladesh has increased its tea supply. According to the BTB figures, over 96.51 million kg were produced in 2021, an increase of almost 54% from that of 2012. As tea is a profitable crop from Bangladesh’s perspective, tea yield prediction can play a significant role in increasing the production of tea. In this paper, tea yield has been forecast using various machine learning algorithms based on area-based tea production and their weather data from 1968 to 2021. Necessary data has been collected from BBS, BARC, and BMD government organizations. Eight climate factors have been used for the research. The data model has been evaluated by using eight classification and regression algorithms, with the Random Forest classifier showing the best accuracy of 97%. With this approach, tea production can be increased while reducing food threats and managing it for other nations.