论文部分内容阅读
针对卫星地面站资源配置问题输入输出关系的复杂性,提出支持向量机(SVM)回归模型。在模型学习过程中,利用正交设计抽样抽取小数量样本对SVM进行训练,求解二次规划问题,根据测试结果对模型参数进行调整以实现较小的预测误差。实验结果表明,该方法具有较好的拟合精度和泛化能力。
In view of the complexity of input and output relationship of satellite ground station resource allocation, a support vector machine (SVM) regression model is proposed. In the process of model learning, a small number of samples are sampled by orthogonal design to train the SVM, quadratic programming problems are solved, and the model parameters are adjusted according to the test results to achieve smaller prediction errors. Experimental results show that this method has better fitting accuracy and generalization ability.