Predicting Crop Prices using Machine Learning Algorithms for Sustainable Agriculture

The agricultural sector is crucial for South Asia's economy, providing employment and driving GDP growth. Accurate price forecasting is essential for food security, market stability, and facilitating import-export activities. In this study, Decision Tree, AdaBoost, and Random Forest have proven effectiveness in predicting crop prices after analyzing different regression models. The Decision Tree has achieved R-squared <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(R^{2})$</tex> values of 0.996, 0.996, and 1 and Mean Squared Error MSE values of 0.242, 4.827, and 0.846 respectively for Rice, Onion, and Oil Seed. Similarly, the AdaBoost has also performed well with R-squared <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(R^{2})$</tex> values of 0.737, 0.959 and 1 and MSE values of 175.234, 14.566 and 1.643 respectively for Tomato, Potato and Oil Seed, indicating better prediction accuracy and lower prediction error. Lastly, Wheat has performed well with Linear Regression with R-squared <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(R^{2})$</tex> values of 0.980 and MSE values of 0.756. Random Forest Regression has shown consistent high R-squared <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(R^{2})$</tex> values and low MSE values. The effectiveness of Decision Tree, AdaBoost, and Random Forest in Bangladesh's agricultural price forecasting highlights their versatility, aiding informed decisions by farmers and policymakers.