Study And Forecast Oil Freight Based On Neural Network

Oil Freight Index is one of the important economic indexes in oil shipping market. Because of its constantly changing and many factors influencing it, most interesting topic is how to grasp accurately Oil Freight Index in oil shipping market. It is considered as barometer reflecting oil shipping market's variation. Importance of studying fluctuant system of Oil Freight Index and inquiring into its variant tendency has been attached more and more.The dissertation analyzes and forecasts Oil Freight Index from qualitative and quantitative view. Analyze factors influencing Oil Freight Index from qualitative view. The main factors are supply and demand. Oil shipping market supply is concerned about shipbuilding and fleet while demand is concerned about oil inventories, oil demand, oil product and seaborne imports& exports. In addition, the other factors such as going on strike, horrible action, war etc. are reflected in the dissertation.Forecast variable Oil Freight Index using BP neural network from quantitative view. First make original data processed. Then establish a reasonable model according to neural network theory and requirement and put processed data into the model. Train repeatedly until the result is super. The method can handle not only short-term forecasting but also long-term forecasting which makes up traditional forecasting way ——time series model. Certainly the author forecast original data again using time series model in order to contrast and make a conclusion that the method using BP neural network is better.