Access Journal

The study is devoted to analysing the economic role of one of the main factors of agribusiness - soil fertility. It is of great importance in the survival of ecosystems and directly affects crop productivity, which determines a country’s GDP. The economic assessment of land in agribusiness is undoubtedly associated with challenges such as predicting soil fertility, which largely correlates with the attractiveness of investment for certain start-ups in agribusiness. This study proposes a simple and understandable model for the economic assessment of land productivity based on an analysis of modern scientific literature and expert opinions. Methods and results: this research proposes a generative model that utilises the collected data to generate additional soil sample data. The produced data are manually reviewed by an expert and validated using a Student’s t-hypothesis test. These data are labelled using a variety of machine learning classifiers. We use SHAP and LIME to interpret the performance of the machine learning classifiers. The effects of various soil constituents on fertility are thoroughly examined. A thorough discussion of each feature’s contribution to the model’s prediction is also included. Additionally, the features’ correlation is discussed. The interpretations presented in this study are transformed into practical ideas for the agribusiness sector through economic efficiency taking into account discounted future income. In addition, the model proposed in this article indirectly contributes to the balance of investment and profit in agribusiness start-ups. The study’s results may be helpful to those conducting economic valuation of land and investment funds, individuals planning agribusiness start-ups, and researchers in the related fields of the study.