Análise Baseada em Aprendizado de Máquina dos Determinantes de Subsistência em Sistemas Agrícolas Peruanos: Uma Abordagem Multirregional Usando Random Forest e Gradient Boosting
DOI:
https://doi.org/10.17268/sci.agropecu.2026.47Palavras-chave:
agricultura de subsistência, aprendizado de máquina, Random Forest, pequenos agricultores, comercialização agrícola, PeruResumo
A agricultura de subsistência permanece como desafio crítico para desenvolvimento rural no Peru, com padrões heterogêneos entre regiões ecológicas. Compreender determinantes dos níveis de subsistência é essencial para intervenções políticas direcionadas, mas análises multirregionais usando métodos preditivos avançados são escassas. Este estudo analisou determinantes de subsistência familiar em sistemas agrícolas peruanos usando modelos de machine learning (Random Forest e Gradient Boosting), identificando padrões regionais e fatores-chave em Costa, Andes e Amazônia. Foram analisados dados da Pesquisa Nacional Agropecuária 2024 do Peru (ENA 2024, INEI), com 34.827 unidades agropecuárias após filtros de qualidade. Subsistência foi operacionalizada como percentagem de autoconsumo sobre valor total de produção, categorizada como Baixa (0-20%), Média (20-60%) e Alta (>60%). Ambos modelos alcançaram precisão convergente (RF: 82,92%, GB: 82,79%), com excepcional estabilidade na validação cruzada (DP<0,4%). Renda per capita emergiu como preditor dominante (importância média=33,5%), seguida de renda total (26,1%) e margem agrícola (7,5%). Disparidades regionais foram extremas: agricultores costeiros ganharam 31,9× mais e retiveram margens 63,9× maiores que andinos, apesar de sistemas andinos exibirem 35% maior diversificação produtiva. Análise de canais mostrou que venda direta ao consumidor associou-se com maior subsistência (45%) versus intermediação por atravessadores (20%), indicando barreiras de acesso a mercado mais que benefícios de desintermediação. Capacidade econômica determina esmagadoramente níveis de subsistência sobre estratégias de diversificação produtiva, contradizendo paradigmas tradicionais de desenvolvimento agrícola. Heterogeneidade regional requer abordagens políticas diferenciadas, com métodos de machine learning capturando efetivamente relações não lineares complexas e especificidades regionais.
Referências
Abdullahi, M. A., Muse, A. H., Muse, Y. H., Hassan, M. A., & Nadarajah, S. (2026). Harvests under heat: Quantifying the long-run impacts of climate and fertilizers on Somalia cereal yields: An ARDL analysis. Earth Systems and Environment, 10(5), 6599-6610. https://doi.org/10.1007/s41748-025-00972-w
Akber, M. A., Islam, M. A., Rahman, M. M., & Rahman, M. R. (2022). Crop diversification in southwest coastal Bangladesh: Insights into farming adaptation. Agroecology and Sustainable Food Systems, 46(2), 316-324. https://doi.org/10.1080/21683565.2021.1989105
Aniye, H. W., Bekele, T., & Worku, W. (2025). Trends of climate variability and its association with farmers' perceptions of climate change in Gassera district, southeastern Ethiopia. Scientific Reports, 15(1), Article 4399. https://doi.org/10.1038/s41598-025-89017-w
Asamoah, E., Heuvelink, G. B. M., Chairi, I., Bindraban, P. S., & Logah, V. (2024). Random forest machine learning for maize yield and agronomic efficiency prediction in Ghana. Heliyon, 10(17), e37065. https://doi.org/10.1016/j.heliyon.2024.e37065
Blanco, M., Mosquera, L. E., Crisostomo, D., Díaz Casana, C., Ritter, T., & Mockshell, J. (2024). Exploring cacao business models and agroecological transitions in Ucayali, Peru (Agroecological Transitions Program Report). Bioversity International and International Center for Tropical Agriculture.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5-32. https://doi.org/10.1023/A:1010933404324
Calderan Gregolin, A., Contreras, M., Salgado-Funes, E., Tejada, G., & Zabaleta, I. (2023). The use of efficient Smart Agro 4.0 irrigation systems for cotton in Peru and its public policy implications. World Water Policy, 9(4), 746-755. https://doi.org/10.1002/wwp2.12157
Coelho da Silva, T. L., Furtado, L. S., Fernandes, G. S., Fernandes de Macêdo, J. A., Almada Cruz, L., Pires Magalhães, R., & Amorim Gomes, L. G. (2025). Machine learning methods and models to predict food insecurity levels for families in Ceará, Brazil, based on employment, housing and other social indicators. Journal of the Brazilian Computer Society, 31(1), 203-218. https://doi.org/10.5753/jbcs.2025.3100
Cole, R. (2022). Cashing in or driving development? Cross-border traders and maize contract farming in northeast Laos. Journal of Agrarian Change, 22(1), 139-161. https://doi.org/10.1111/joac.12460
Donkor, E. A., Garnevska, E., Siddique, M. I., & Donk or, E. (2021). Determinants of rice farmer participation in the direct marketing channel in Ghana. Sustainability, 13(9), 5047. https://doi.org/10.3390/su13095047
Food and Agriculture Organization of the United Nations. (2025). The state of food and agriculture 2025: Addressing land degradation across landholding scales. FAO. Rome.
Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. Annals of Statistics, 29(5), 1189-1232. https://doi.org/10.1214/aos/1013203451
Fuglie, K. O. (2018). Is agricultural productivity slowing? Global Food Security, 17, 73-83. https://doi.org/10.1016/j.gfs.2018.05.001
Gholami, S., Knippenberg, E., Campbell, J., Andriantsimba, D., Kamle, A., Parthasarathy, P., Sankar, R., Birge, C., & Lavista Ferres, J. (2022). Food security analysis and forecasting: A machine learning case study in southern Malawi. Data & Policy, 4, e33. https://doi.org/10.1017/dap.2022.25
Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning (2nd ed.). Springer. https://doi.org/10.1007/978-0-387-84858-7
Instituto Geográfico Nacional. (2023). Límites departamentales [Data set]. Plataforma Nacional de Datos Abiertos. https://www.datosabiertos.gob.pe/dataset/limites-departamentales
Instituto Nacional de Estadística e Informática. (2024). Encuesta Nacional Agropecuaria 2024: Ficha técnica. Dirección Nacional de Censos y Encuestas, INEI. https://proyectos.inei.gob.pe/iinei/srienaho/Descarga/DocumentosMetodologicos/2024-62/03_FICHA_TECNICA_ENA_2024.pdf
James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An introduction to statistical learning with applications in R. Springer. https://doi.org/10.1007/978-1-4614-7138-7
Jebessa, G. M., Engura, T. T., & Bekele, A. E. (2026). Exploring the nexus between forest resource use, livelihood strategies, and household food security in Ethiopia. Discover Food, 6(1), Article 285. https://doi.org/10.1007/s44187-026-01021-9
Jimoh, A. K., Olawuyi, S. O., Balogun, T. A., & Ganiyu, M. O. (2024). Causal effect of agricultural diversification on smallholder farmers' welfare status in Nigeria. Research on World Agricultural Economy, 5(4), 458-480. https://doi.org/10.36956/rwae.v5i4.1223
Leah, J., Pradel, W., Cole, D. C., Prain, G., Creed-Kanashiro, H., & Carrasco, M. V. (2013). Determinants of household food access among small farmers in the Andes: Examining the path. Public Health Nutrition, 16(1), 136-145. https://doi.org/10.1017/S1368980012000183
Little, R. J. A., & Rubin, D. B. (2019). Statistical analysis with missing data (3rd ed.). Wiley. https://doi.org/10.1002/9781119482260
Louppe, G., Wehenkel, L., Sutera, A., & Geurts, P. (2013). Understanding variable importances in forests of randomized trees. Advances in Neural Information Processing Systems, 26, 431-439.
Lowder, S. K., Skoet, J., & Raney, T. (2016). The number, size, and distribution of farms, smallholder farms, and family farms worldwide. World Development, 87, 16-29. https://doi.org/10.1016/j.worlddev.2015.10.041
Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765-4774.
Ma, W., Rahut, D. B., Sonobe, T., & Gong, B. (2024). Linking farmers to markets: Barriers, solutions, and policy options. Economic Analysis and Policy, 82, 1102-1112. https://doi.org/10.1016/j.eap.2024.05.005
Mazenda, A. (2026). Predicting household food insecurity through a livelihood lens: A machine learning approach in rural Colombia. Journal of International Development, 38(5), 827-840. https://doi.org/10.1002/jid.70092
Mgomezulu, W. R., Chitete, M. M. N., Maonga, B. B., Kachingwe, L., Phiri, H. H., Mambosasa, M., & Folias, L. (2024). Does shifting from subsistence to commercial farming improve household nutrition and poverty? Evidence from Malawi, Tanzania and Nigeria. Research in Globalization, 8, 100201. https://doi.org/10.1016/j.resglo.2024.100201
Ministerio de Desarrollo Agrario y Riego. (2021). Metodología de la construcción del marco maestro muestral para la ENA-MIDAGRI y demás operaciones estadísticas en el periodo intercensal (1.ª ed.). Unidad Ejecutora Gestión de Proyectos Sectoriales, Proyecto PIADER. https://ena.midagri.gob.pe/media/attachments/2021/10/08/metodologia_construccion_mmm_ena-midagri.pdf
Mukashov, A., Robinson, S., Arndt, C., Thurlow, J., & Thomas, T. (2026). Systematic risk profiling: Assessing compounding economic risks in developing countries. Economic Modelling, 157, 107511. https://doi.org/10.1016/j.econmod.2026.107511
Nkurunziza, F., Kabanda, R., & McSharry, P. (2025). Enhancing poverty classification in developing countries through machine learning: A case study of household consumption prediction in Rwanda. Cogent Economics & Finance, 13(1), 2444374. https://doi.org/10.1080/23322039.2024.2444374
Nzira, J. E., & Chegere, M. J. (2025). Diversify and thrive: Panel data evidence on crop portfolio effects on farmer income and farm output market supply. Journal of Agriculture and Food Research, 24, 102339. https://doi.org/10.1016/j.jafr.2025.102339
Ogutu, S. O., Gödecke, T., & Qaim, M. (2020). Agricultural commercialisation and nutrition in smallholder farm households. Journal of Agricultural Economics, 71(2), 534-555. https://doi.org/10.1111/1477-9552.12359
Probst, P., Wright, M. N., & Boulesteix, A.-L. (2019). Hyperparameters and tuning strategies for random forest. WIREs Data Mining and Knowledge Discovery, 9(3), e1301. https://doi.org/10.1002/widm.1301
Solano-Jiménez, J., Abadía, R., Pinilla Palleja, R., & Martínez-Carrasco, L. (2026). Determinants of household food security in rural Uganda: Perceived needs as key predictors. Agriculture & Food Security, 15(1), Article 14. https://doi.org/10.1186/s40066-025-00594-5
Tesfaye, W., & Tirivayi, N. (2020). Crop diversity, household welfare and consumption smoothing under risk: Evidence from rural Uganda. World Development, 125, 104686. https://doi.org/10.1016/j.worlddev.2019.104686
Trivelli, C., & Berdegué, J. A. (2019). Rural transformation: Looking towards the future of Latin America and the Caribbean (2030 - Food, Agriculture and Rural Development in Latin America and the Caribbean No. 1). Food and Agriculture Organization of the United Nations. https://openknowledge.fao.org/handle/20.500.14283/ca5508en
Vallejo-Rojas, V., Rivera-Ferre, M. G., & Ravera, F. (2022). The agri-food system (re)configuration: The case study of an agroecological network in the Ecuadorian Andes. Agriculture and Human Values, 39(4), 1301-1327. https://doi.org/10.1007/s10460-022-10318-1
van Klompenburg, T., Kassahun, A., & Catal, C. (2020). Crop yield prediction using machine learning: A systematic literature review. Computers and Electronics in Agriculture, 177, 105709. https://doi.org/10.1016/j.compag.2020.105709
Vásquez Neyra, J. M., Gomes Haensel Schmitt, V., & Cequea, M. M. (2025). Sustainable circular strategies and short supply chains among smallholder producers: The case of Peru's Agroferias Campesinas. Agricultural and Food Economics, 13(1), 12. https://doi.org/10.1186/s40100-025-00423-0
Wilson, K. M., McCool, W. C., & Brenner Coltrain, J. (2024). Climate and oceanic condition changes influence subsistence economic adaptation through intensification on the central Andean coasts. Quaternary International, 689-690, 79-92. https://doi.org/10.1016/j.quaint.2023.07.002
Zantsi, S., Masola, L., & Shiba, W. (2026). Appropriate farmer classification is key to agricultural development: Rethinking subsistence agriculture in South Africa. International Journal of Social Economics, 53(8), 1288-1298. https://doi.org/10.1108/IJSE-04-2025-0401
Zhao, L., Yang, M., Min, S., & Qing, P. (2025). Prediction of household food insecurity in rural China: An application of machine learning methods. International Food and Agribusiness Management Review, 28(2), 296-316. https://doi.org/10.22434/ifamr1137
Downloads
Publicado
Edição
Seção
Licença
Direitos autorais (c) 2026 Scientia Agropecuaria
Este trabalho está licenciado sob uma licença Creative Commons Attribution-NonCommercial 4.0 International License.
The authors who publish in this journal accept the following conditions:
a. The authors retain the copyright and assign to the magazine the right of the first publication, with the work registered with the Creative Commons attribution license, which allows third parties to use the published information whenever they mention the authorship of the work and the First publication in this journal.
b. Authors may make other independent and additional contractual arrangements for non-exclusive distribution of the version of the article published in this journal (eg, include it in an institutional repository or publish it in a book) as long as it clearly indicates that the work Was first published in this journal.
c. Authors are encouraged to publish their work on the Internet (for example, on institutional or personal pages) before and during the review and publication process, as it can lead to productive exchanges and a greater and faster dissemination of work Published (see The Effect of Open Access).
