The Research On The ACO-SVM Real Estate Investment Risk Evaluation

The real estate industry is a fundamental and pioneering industry in our national economy. The real estate investment is a high investment, high risk, high return in investment activities. It affected by many factors, such as economic, social, technical, natural and managers’ quality. Due to the rapid booming of the real estate market in recent years, many investors even if only by subjective experience rough invest decision, also can obtain higher profits, thus ignore the research and analysis of investment risk factors. Along with our country’s housing situation from shortage to basic balance, and even local demand slightly greater than the supply situation,once the risk occurs, investors will bring irreparable loss, if don’t make scientific analysis to real estate project investment risk. Therefore, it has important practical significance to make comprehensive analysis and scientific evaluation on the real estate project investment risk.This paper first studied the current situation of the real estate investment, analysis of real estate investment risk research progress at home and abroad. Secondly, basing on the theory of real estate investment risk, according to the source of the risk factors, which can be divided into social factors, economic factors, technical factors, natural and internal management risk 19 fall into 6 categories, specific risk factors, and build the risk evaluation index system. Then, select least squares support vector machine(SVM), which have a good effect in forecasting area, as the main method of risk evaluation, and use the ant colony algorithm(ACO) to overcome blindness of SVM in the parameter choice, based ACO-SVM is constructed risk assessment model. Finally, comparing the ACO-SVM model with the cross validation support vector machine mode by using the MATLAB programming with 25 investment project dates in recent 3 years, then verify the validity of the ACO-SVM model.The model, which basing on the ACO- SVM model in this paper, used to predict the risk of real estate investment, having important practical significance to improve the ability of investors’ risk analysis and investment decision.