A new BDI forecasting model based on support vector machine

BDI (Dry Baltic Index) is a barometer of the shipping market, which is very important and difficult to predict precisely. The traditional forecasting method is not enough to predict precisely for it with higher complexity. The support vector machine has strong nonlinear function approximation and strong generalization ability. In this paper, a new BDI forecasting model based on support vector machine combined with CFS(Correlation-based Feature Selection) is presented and we first introduce the macroeconomic fundamental indicators of BDI. Firstly, we studied and researched the main macroeconomic fundamental indicators of BDI, investigate the inherent law and external influence of freight index fluctuation, and then put forward the BDI forecasting model based on support vector machine and demonstrate the construction principle of the forecasting model. Finally, in order to evaluate the predictive power of support vector machines, we compare their performance with the neural network model. The results show that the SVM model has a better performance in both the trend and the forecast precision, which can provide a powerful tool for the shipping market operators and investors to grasp the market trend and avoid the price risk.