Data Mining Techniques Applied to the Analysis and Prediction of Sugarcane Yield

• Grant number: 12/50049-3
• Support Opportunities: Program for Research on Bioenergy (BIOEN) - Research Partnership for Technological Innovation (PITE)
• Start date: April 01, 2013
• End date: March 31, 2015
• Field of knowledge: Agronomical Sciences - Agricultural Engineering
• Agreement: Odebrecht Agroindustrial
• Principal Investigator: Luiz Henrique Antunes Rodrigues
• Grantee: Luiz Henrique Antunes Rodrigues
• Host Institution: Faculdade de Engenharia Agrícola (FEAGRI). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Company: ETH Bioenergia
• City: Campinas

Abstract

For the agricultural planning of sugarcane crop, it is necessary to understand the relationship between the crop yield and the production environment, management performed, climate, and variety and maturity of the plant. Currently, the most widely used yield estimates are subjective; experts perform this work based on their experience and field evaluations. There are two crop yield predictive models: ecophysiological, based on the physiological processes, and empirical-statistical, where analysis of past climate variables and crop yield are made, by using traditional statistical methods. Ecophysiological models are seldom used in the plant level, due to the complexity of data required and lack of detailed knowledge of physiology; empirical-statistical models have not been proven to be the best prediction tool for dealing with such complex data. Among these models, there is no consensus or reason concerning the variables that should be used or the level of detail necessary to achieve better performance. Other difficulty in this industry is to understand the significance of the influence of the interactions of production factors on crop yield. This knowledge may be incorporated in the extensive data bases of the plants. Hence, the use of appropriate tools for their extraction is suitable. Data Mining techniques were popularized due to their ability to find intelligible knowledge in databases and by their potential to predict, including forecast applications in agriculture. In this paper, we propose to use Data Mining techniques to investigate the variability of sugarcane crop yield and determine the best set of variables to accomplish this task; we also propose to use these techniques to find models that provide better sugarcane yield predictions. (AU)

Scientific publications

(6)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

PELOIA, PAULO RODRIGUES; BOCCAE, FELIPE FERREIRA; ANTUNES RODRIGUES, LUIZ HENRIQUE. Identification of patterns for increasing production with decision trees in sugarcane mill data. Scientia Agricola, v. 76, n. 4, p. 281-289, JUL-AUG 2019. (12/50049-3)

BOCCA, FELIPE F.; ANTUNES RODRIGUES, LUIZ HENRIQUE. The effect of tuning, feature engineering, and feature selection in data mining applied to rainfed sugarcane yield modelling. COMPUTERS AND ELECTRONICS IN AGRICULTURE, v. 128, p. 67-76, OCT 2016. (12/50049-3)

GRAVINA DE OLIVEIRA, MONIQUE PIRES; BOCCA, FELIPE FERREIRA; ANTUNES RODRIGUES, LUIZ HENRIQUE. From spreadsheets to sugar content modeling: A data mining approach. COMPUTERS AND ELECTRONICS IN AGRICULTURE, v. 132, p. 14-20, JAN 2017. (12/50049-3)

PELOIA, PAULO RODRIGUES; BOCCAE, FELIPE FERREIRA; ANTUNES RODRIGUES, LUIZ HENRIQUE. Identification of patterns for increasing production with decision trees in sugarcane mill data. CIENTIA AGRICOL, v. 76, n. 4, p. 9-pg., 2019-07-01. (12/50049-3)

BOCCA, FELIPE FERREIRA; ANTUNES RODRIGUES, LUIZ HENRIQUE; MODESTO ARRAES, NILSON ANTONIO. When do I want to know and why? Different demands on sugarcane yield predictions. AGRICULTURAL SYSTEMS, v. 135, p. 48-56, MAY 2015. (12/50049-3)

PELOIA, PAULO R.; RODRIGUES, LUIZ H. A.. IDENTIFICATION OF COMMERCIAL BLOCKS OF OUTSTANDING PERFORMANCE OF SUGARCANE USING DATA MINING. Engenharia Agrícola, v. 36, n. 5, p. 895-901, SEP-OCT 2016. (12/50049-3)