Resumo

RUIZ-RAMIREZ, Juan; HERNANDEZ-RODRIGUEZ, Gabriela Eréndira  e  ZULUETA-RODRIGUEZ, Ramón. Time series analysis forecasting in sugar cane production. Terra Latinoam [online]. 2011, vol.29, n.1, pp.103-109. ISSN 2395-8030.

Economic entities generally need to implement methods to forecast their performance as accurately as possible to stay on top in the competitive market. Thus, decision-making in the mexican sugar industry needs to predict the behavior of its productive process in order to plan and optimize the use of the technical, financial and human resources. To improve the profitability of the sugar cane refineries, time series data sets are used as they allow prediction of the expected returns and thus determine future demands on the inputs and inventories. In this research we used a time series model to forecast the Independencia sugar cane mill 2006-2007 production. The statistica software was used to analyze the volume of harvests from 1949 to 2006, and with the Box-Jenkins methodology to generate an auto-regressive integrated moving average (ARIMA) model. The 2006-2007 sugarcane cycle production forecasted 11 974 tons, in contrast with the obtained production (12 736 tons). Our results indicate that forecasting with the model is 94% accurate. This result was influenced by reducing planting area by 1.5%.

Palavras-chave : time series model; sugarcane harvest; ARIMA model.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )