Predicting price of target in acquisition by support vector machine

In order to predict the price of candidates in acquisition and evaluate its feasibility, this paper puts forward a model of price prediction of candidates based on support vector machine. The model is trained by the data of market deals which were made in the past. The result of simulation and test indicates that average error of prediction is percent 7.71. It also proves that SVM has better performance than BP neural network under the condition of small samples. SVM can overcome the limitation of hypotheses of econometrics and shortcomings of BP neural network that is low efficient and easy to fall into local minimum. Therefore, prediction model based on SVM has more advantages and wider prospect of application.