Housing price has always been one of the hot issues concerned by scholars at home and abroad as well as the people.The price of house hits record highs having been widely concerned by people,if allowed to develop,it will affect the stability of society and macro economy.Prices of houses and people's life can not be avoided,whether from meeting the basic needs of view or from the angle of economic development,knowing the development and change of the trend of price is very important.Therefore the scientific forecast on housing price is very important.From the national level,we can make the policy of market development according to the forecast results.We also can effectively guide the developers to make reasonable prices,so that their own interests are guaranteed.But there are many factors that affect the price,so forecasting price is an extremely complex problem,and the price of data is relatively few,so the application of the model to predict prices is not very numerous.Because the support vector machine has a good effect in the prediction of small samples.In order to achieve a certain prediction accuracy,this paper proposes that using support vector regression to analyze the house prices of Wuhan,and by using particle swarm algorithm to find the most suitable parameters of the model,then using the results of particle swarm optimization algorithm for support vector regression to forecast the housing price of Wuhan.The article mainly analyzes of the related factors affecting the price of the study in terms of theory,and introduces the related concept of commodity residential housing,then finds out the influencing factors of the price according to the actual situation.Then this article introduces the principle of support vector regression.Next,we take commercial residential housing of Wuhan as the research object to analyze.Firstly,the profile of Wuhan and the real estate situation are briefly introduced,and then combined with the front of the factors affecting the price of selected data,using principal component analysis to reduce the dimension of the data.The principle of particle swarm optimization and genetic algorithm is introduced,then,with data from 1998 to 2012 as a test set,using the particle swarm algorithm and genetic algorithm for optimization of the contrast parameters.The particle swarm optimization has better effect in Wuhan City Housing test set,and use it to support vector regression predicting the housing price of Wuhan from 2013 to 2015.In order to further illustrate the superiority of the model,we select multiple linear regression model and BP neural network model forecasting the housing price of Wuhan from 2013 to 2015,the results show that for small samples,the support vector regression has higher effect of prediction than the BP neural network model and multiple linear regression. |