Finding Optimal Farming Practices to Increase Crop Yield Through Global-Best Harmony Search and Predictive Models, a Data-Driven Approach
Increasing crops’ yields to meet the world’s demand for food is one of the great challenges of these times. To achieve this, farmers must make the best decisions based on the resources available for them. In this paper, we propose the use of Global-best Harmony Search (GHS) to find the optimal farming practices and increase the yields according to the local climate and soil characteristics, following the principles of site-specific agriculture. We propose to build an aptitude function based on a random forest model trained on farms’ data combined with open data sources for climate and soil. The result is an optimizer that uses a data-driven approach and generates information on the optimized farming practices, allowing the farmer to harness the full potential of his land. The approach was tested on a case-study on maize in the state of Chiapas, Mexico, where the adoption of the practices suggested by our approach was estimated to increase average yield by 1.7 ton/ha, contributing to closing the yield gap. The proposal has the potential to be scaled to other locations, other response variables and other crops.
KeywordsGlobal-best harmony search Machine learning Open data Data-driven agronomy Optimization
The research has been supported by International Center for Tropical Agriculture (CIAT) and is based on data shared by the MASAGRO project lead by the International Maize and Wheat Improvement Center (CIMMYT), we also acknowledge for the open data shared by INIFAP and INEGI. We are especially grateful to Colin McLachlan for suggestions relating to the text in English.
- 2.Zader, A.: Food and agriculture. In: Routledge Handbook of Environment and Society in Asia, p. 237 (2014)Google Scholar
- 6.Reid, W.V., Berkes, F., Wilbanks, T.J., Capistrano, D., et al.: Bridging Scales and Knowledge Systems: Concepts and Applications in Ecosystem Assessment. Island Press, Washington (2006)Google Scholar
- 8.Samberg, L.H., Gerber, J.S., Ramankutty, N., Herrero, M., West, P.C.: Subnational distribution of average farm size and smallholder contributions to global food production. Environ. Res. Lett. 11(12), 1–12 (2016). http://iopscience.iop.org/article/10.1088/1748-9326/11/12/124010/metaCrossRefGoogle Scholar
- 9.De Janvry, A., Sadoulet, E., Suri, T.: Field experiments in developing country agriculture. In: Handbook of Economic Field Experiments, vol. 2, pp. 427–466. Elsevier, Amsterdam (2017)Google Scholar
- 10.Leeuwis, C.: Communication for Rural Innovation: Rethinking Agricultural Extension. Wiley, Hoboken (2013)Google Scholar
- 12.Jiménez, D., et al.: From observation to information: data-driven understanding of on farm yield variation. PLoS One 11(3), 1–20 (2016)Google Scholar
- 14.Jiménez, D., et al.: Analysis of Andean blackberry (Rubus glaucus) production models obtained by means of artificial neural networks exploiting information collected by small-scale growers in Colombia and publicly available meteorological data. Comput. Electron. Agric. 69(2), 198–208 (2009)CrossRefGoogle Scholar
- 16.Strobl, C., Boulesteix, A., Kneib, T., Augustin, T., Zeileis, A.: Conditional variable importance for random forests. BMC Bioinform. 11, 1–11 (2008)Google Scholar
- 22.Kennedy, J., Eberhart, R.: Particle swarm optimization. In: Proceedings of IEEE International Conference on Neural Networks, vol. 4, pp. 1942–1948 (1995)Google Scholar
- 27.Hagan, M.T., Demuth, H.B., Beale, M.H., De Jesús, O., et al.: Neural Network Design, vol. 20. Pws Pub, Boston (1996)Google Scholar
- 36.Assad, A., Deep, K.: Applications of harmony search algorithm in data mining: a survey. In: Pant, M., Deep, K., Bansal, J., Nagar, A., Das, K. (eds.) Proceedings of Fifth International Conference on Soft Computing for Problem Solving. Advances in Intelligent Systems and Computing, vol. 437, pp. 863–874. Springer, Singapore (2016). https://doi.org/10.1007/978-981-10-0451-3_77CrossRefGoogle Scholar
- 37.Kuhn, M., et al.: Caret: Classification and Regression Training (2014)Google Scholar