Study on extended Gray Model for sales prediction in automotive market with Scarce Data

摘要

Recently, Product lifecycle becomes short and many new brand products are created frequently, especially in Chinese automobile market. Therefore, for an automobile enterprise, they need to make a decision how many new brand cars they need to produce during the different periods. However, we may not have enough historical sales data of new brand car, therefore conventional forecasting method doesn't work well. So we proposed Gray forecasting model which require small number of data in this research. And we also proposed genetic algorithm and seasonal index model to improve the performance of Gray forecasting model based on its shortcomings.