CAJ | 학술논문

在保证足够信息量的前提下,针对合理减少气象观测站的实际问题,首先利用主成分分析( PCA)降低样本数据的维数.其次利用支持向量回归机(SVR)对样本进行有效的回归.然后结合优化软件lingo对凸二次规划问题(与支持向量回归机相对应)进行求解.最终得出基于主成分分析,支持向量机回归预测优化模型.
재보증족구신식량적전제하,침대합리감소기상관측참적실제문제,수선이용주성분분석( PCA)강저양본수거적유수.기차이용지지향량회귀궤(SVR)대양본진행유효적회귀.연후결합우화연건lingo대철이차규화문제(여지지향량회귀궤상대응)진행구해.최종득출기우주성분분석,지지향량궤회귀예측우화모형.
In the premise to ensure enough information,some measures are taken to reduce meteorological observation station properly.Dimensions of samples are reduced with the principal component analysis method,then samples are regressed effectively based on support vector regression ( SVR) finally the regressive forecast model is established combining with Linear Interactive and General Optimizer to solve the convex quadratic programming corresponding to support vector regression.