In this thesis, the researcher studied the mathematical relationship. To estimate the rate of loss of sugarcane production in each season. And each process using sugar. Approximation Theory, whose main objective is to study the process of production of the sugar industry. Analysis is performed in two styles. 1) analyze the rate of loss resulting from the production of sugarcane in the production of sugar, 2) analyze the rate of loss resulting from the production of sugarcane at various times of the production season. It also compares and analyzes the function of estimation of loss resulting from the production of sugarcane by the least squares method in a Linear Least Squares, Quadratic Least Squares and Exponential Least Squares. With all the preview assembly using a mathematical program to use to calculate the coefficients of approximation function performance. Linear Least Squares method showed that charges the least tolerance is 0.9406 suit case study because the trend rate of economic losses from the production process of continuous steady decline every year because of plant development, improve and enhance the performance continuously.