基于支持向量机的多维度数据融合在油气预测中应用磁

摘要

There are some insurable problems in prediction of oil and gas ,including the selection of attribute ,the selec-tion of the algorithm etc .To do best to reduce the phenomenon that there are some insurable problems ,this paper introduces the means that is blending with SVM and information fusion to predict the oil and gas .By using the transformation by inner production functions of SVM to put the input space to map into a high dimensional space ,the measure that is blending with SVM and information fusion to predict the oil and gas to reduce the insurability to enhance the accuracy of oil and gas is taken ,and the experiment is taken to verify the results .%油气预测中存在诸多不确定因素，包括属性的选取、预测算法的选取等。为了尽量消除这些不确定因素，论文利用支持向量机与信息融合理论进行地震油气预测，通过使用支持向量机内积函数定义的变换将输入空间映射到高维空间，并运用信息融合与支持向量机相结合的方法，尽可能的减小多种因素所带来的不确定性，提高油气预测精度，并通过实际数据验证，得到很好的效果。