Estimating and optimizing the productive of different species of sugar cane using machine learning techniques and genetic algorithms

• Grant number: 17/01458-1
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: July 01, 2017
• End date: June 30, 2018
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: André Luis Debiaso Rossi
• Grantee: Víctor Augusto Faria Martins
• Host Institution: Universidade Estadual Paulista (UNESP). Campus Experimental de Itapeva. Itapeva , SP, Brazil

Abstract

Brazil is one of the world's major producers of alcohol and the world's largest exporter of sugar. Moreover, sugarcane is one of the largest and most important sources of energy for the country and great relevance for its economy. The Midwest, Southeast and Southern regions comprise the largest portion of sugarcane produced in Brazil, being that state of São Paulo is the largest producer in the country. However, the same can not be said of its productivity within the regions mentioned, since in the last years the exploitation of the planted area was below its potential. In addition, the estimated productivity of sugarcane is declining for the next years, making valid and relevant studies that aim to analyze which aspects influence the productivity of sugarcane and how to improve the use of its planting. Considering the current scenario, this research project has the objective of optimizing the planting of different sugarcane varieties using estimates of the proportion of fiber, sucrose and productivity. These estimates will be generated by Machine Learning techniques, which have been used successfully in extracting patterns into specific problems for creating descriptive and predictive models. The optimization will be performed by Genetic Algorithms metaheuristics, which is based on ideas from the evolutionary process to find the best combination of the main factors that influence the productivity of sugarcane in a given region. For this purpose, data from various sugarcane varieties planted at more than 300 farms in the Midwest, Southeast and Southern regions of Brazil will be analyzed. (AU)