中文摘要：

为分析预测干散货航运市场运价波动的警情,建立基于支持向量机的运价预警模型,并构造相应的算法。选择波罗的海海岬型指数（Baltic Capesige Index BCI）、波罗的海巴拿马指数（Baltic Panamax Index BPI）、波罗的海灵便型指数（Baltic Supramax Index BSI）、波罗的海小灵便型指数（Baltic Handsige Index BHSI）等四个干散货运价指数作为警兆指标,结合航运专家知识经验,确定干散货航运市场运价的实际警度。依据训练样本数据,利用支持向量机的学习功能,通过编制MATLAB软件程序,获得市场运价警度的分类超平面及预测警度区间,并进行内插和外推检验。检验结果表明此方法对于干散货航运市场运价预警有很好的适用性。

英文摘要：

A pre-warning model of freight rate based on support vector machine is established and its algorithm is given to analyze indexes BCI （Baltic Capesige Index）, BPI （Baltic Panamax Index）, BSI （Baltic Supramax Index）, BHSI（Baltic Handsige Index） and forecast the freight rate fluctuation in dry bulk shipping market. Combined with experts＇ knowledge and experiences, the actual alarming degrees of freight rate of dry bulk in shipping market within the sample interval can be determined. The classification super planes and forecasting intervals of freight rate alarming degree are obtained by using the learning function of the support vector machine and MATLAB software, based on training sample data. And the examination of interpolation and extrapolation is also carried out. Examination results show that this methodology has a very good serviceability for pre-warning of freight rate of dry bulk shipping market.