Abstract: Aiming at the product sale time series of manufacturing enterprise with multi-dimension, small samples, nonlinearity, multi-peak, etc., chaotic mapping theory is combined with parameter optimization method of support vector machine, and a kind of chaotic v-support vector machine (SVM) named Cv-SVM is proposed. And then, a product sale forecasting method is put forward. The results of application in car sale forecasting show that the forecasting method based on Cv-SVM is effective and feasible.

Key words: Chaos, Genetic algorithm, Sale forecasting, Support vector machine