Support Vector Machine Prediction Modeling for Automobile Ownership

To address some inherent defects of artificial neural networks, such as insufficient generalization performance, local extremum problem, and dimensional catastrophe problem, a support vector machine was proposed and applied to the modeling of automobile ownership prediction. By analyzing the data on automobile ownership and its influencing factors, the learning sample couples for automobile ownership prediction modeling were constructed, and support vector machine (SVM) was used to regression the nonlinear function relation of automobile ownership prediction model, and the established automobile ownership prediction model was used to predict the automobile ownership in different years. To reduce the impact on the accuracy of automobile ownership prediction caused by the large order of magnitude difference between the data of automobile ownership and its influence factors, the normalization method was used to pre-process the automobile ownership and its influence factors, and the inverse normalization was used to process the automobile ownership prediction results. The comparison between the automobile ownership prediction results and the statistical results shows that the automobile ownership prediction model has good generalization performance, and the support vector machine is an effective method to model the automobile ownership prediction.