Crop yield prediction in Ethiopia using gradient boosting regression

Nowadays, machine learning algorithms and methods are used in multiple areas of studies to achieve practical and productive solutions. Agriculture is one of the industries where the impact issignificant, especially in the area of crop yield prediction and crop selection which is crucial for ensuring food security and improving agricultural practices. In a country like Ethiopia, where the economy is highly dependent on agriculture, and farming in particular, leveraging the powers of AI and machine learning is crucial. However, the use of these technologies in Ethiopian agriculture remains limited, mainly due to the lack of well-organized and digital datasets and lack of technological advancements. The aim of this study is to increase the accuracy of crop yield prediction in Ethiopia and provide information that can help farmers and policymakers improve crop productivity. In this study, a crop yield prediction model was developed based on historical data that includes factors such as crop type, rainfall, temperature, Area cultivated, production, and pesticides. Among the algorithms considered in this study, GradientBoostingRegressor achieved the highest value of the R-square – 90% compared to others which indicates its best predictive ability. However, the study also acknowledges the contextual advantages of other algorithms, highlighting the importance of selecting models that are appropriate for specific data sets and purposes. The accuracy and efficiency of agricultural planning and resource allocation in Ethiopia can be greatly improved by using machine learning techniques for crop production prediction.