RatioLogo
Back

AI Forecasts Crop Prices

New study reveals how machine learning can predict farming costs, offering significant benefits to farmers worldwide.

Food prices are crucial for everyone, from producers to consumers. Historically, predicting these prices has been extremely challenging. However, a new study demonstrates how advanced machine learning algorithms are significantly improving price forecasting accuracy.

The Research Approach

Researchers reviewed 27 scientific papers from major online knowledge repositories. Their focus was on studies utilizing machine learning to predict prices for agricultural crops.

Key Findings

  • Geographical Focus: A significant portion of this research is concentrated in developing countries, including India, China, Bangladesh, Brazil, and Thailand.
  • Most Studied Crops: The most frequently analyzed staple crops include:
    • Corn (8 articles)
    • Rice (7 articles)
    • Wheat (6 articles)
    • Soybeans (5 articles)
  • Common Algorithms: The most common and effective machine learning models employed are Long Short-Term Memory (LSTM) and neural networks (16 articles), which are inspired by the human brain's structure.

The authors shared that, "By integrating these variables, researchers aim to create more comprehensive and robust models for agriculture price prediction." This highlights the importance of incorporating diverse data points beyond historical prices to enhance prediction accuracy.

Implications and Future Directions

Improved price forecasts empower farmers to make more informed decisions regarding planting and selling, potentially reducing losses and ensuring a stable food supply.

However, the study also identified certain limitations:

  • Most research is currently focused on developing countries.
  • There is limited work on predicting prices for groups of crops (e.g., all grains combined).

Future studies could explore broader predictions by grouping similar crops.

Machine learning demonstrates great promise in accurate crop price prediction, contributing to more efficient global food systems.

Reference

Tran, N.-Q., Nguyen, T. N., Tran, Q., Felipe, A., A., Huynh, T., Tang, A., & Nguyen, T. (2023). Predicting Agricultural Commodities Prices with Machine Learning: A Review of Current Research. SOICT 2023, Hochiminh, Vietnam.