Prediction of Cigarette Sales Amount and Average Price based on Lunar Calendar

Prediction of cigarette sales amount and average price is important for anomaly detection of orders, customers and cigarette brands. Besides this, prediction can be support of storing and production arrangement. By analyzing large amount historical sales records, a cigarette sales amount and average price foresting model is built to predict the weekly sales based on ARIMA model which using lunar time as the calendar. Result on the five years sales data from a Chinese city indicate that sales prediction based on ARIMA using lunar time can be an effective way to improve forecasting accuracy.