Influence of Socioeconomic Variables and Skills on the Overall Productivity of a Farm: A Gradient Boosting Based Approach

Abstract

This study analyzes the influence of several variables on the overall productivity of a farm using machine learning techniques. A Gradient Boosting Regressor model was applied to evaluate the relationship between individual characteristics and productivity, using a quantitative approach. The data were divided into a training (80%) and test (20%) set to ensure reproducibility and evaluate the model using metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and the coefficient of determination (R²). The model results indicate that interpersonal skills, age, and education level have a significant impact on productivity, with relative significances of 26.48%, 22.04%, and 20.12%, respectively. The analysis highlights that interpersonal skills are fundamental to optimizing the work environment, while age and education level are correlated with knowledge and experience applied to farm management. On the other hand, variables such as environment, agricultural technician, and administrative accountant did not show a significant influence, probably due to the lack of variability in the data or their lower relevance in the context studied. The findings underscore the need to invest in training and development of interpersonal skills, as well as in continuous training programs to maximize productivity. This study provides data-driven insights that can inform operational strategies and strategic decisions to increase efficiency in farm management. However, limitations were identified in the model, such as low explained variability (R² = 0.22), suggesting the need to incorporate new variables to improve future predictions.