Based On A Combination Of Forecasts Of Revenue Management Research

As the adjustment of the world's and Chinese industry structure, the proportion of the third industry is ascending year by year. Especially the tourism industry, has kept an increase speed of twenty percent since the ninetieth. And as china joined the WTO, the competition became more and more vehement. So we can say that china nowadays is a big stage which has a large number of opportunities and also a lot of threats, on this stage a new thought — Revenue Management appeared. RM is a totally systematical hotel management theory, which is come in to being in the case of the following situations: the competition in the tourism industry is becoming more and more furious and the cost and service are becoming the same. It is a deep mining to the profit of the tourism industry, and it can make the profit become the maximum but don't add fixed cost. From 2000 until now a lot of scholar at home who did the Revenue Management research, but the deep research on the RM's forecast module is very few, this paper's aim is through the construction of six RM models to uncover this black box's dope.The paper starts with RM's history, definition and developing condition, analyzes the differences between airline RM and hotel RM, and puts forward hotel RM's system figure, then opens out hotel RM forecast's module's important. Where after this paper summarizes forecast method, then puts forward this thesis's six hotel RM forecast models: time series forecast model, regression analysis forecast model, neural network one forecast model, neural network two forecast model, combination one forecast model and combination two forecast model.This study has four conclusions and six suggestions. Conclusion:(l)Neural network forecast method is much better than other linear forecast method in the hotel RM area;(2)Importing combination forecast method can make the whole forecast precision of hotel RM much better;(3)Reservation data can enormously affect the regression analysis method and neural network method two;(4)Hotels at home have to make great efforts in the RM. Suggestion:(l)building hotel industry criterion;(2)acquiring high level manager's long support;(3)cultivating a group of high ability RM person;(4)recombining relative department and build RM department;(5)rebuilding performance system;(6) moderate authorization.