摘要
本文在Takens定理基础上,从我国制造业增加值占GDP的比重入手,将微分进化算法和支持向量机有效集成起来,集成联立模型汲取了各方法的优势之处,运用基于微分进化算法的支持向量机模型对我国制造业增加值和制造业增加值占GDP的比重进行预测,集成联立方法有效避免了以往制造业预测研究中多维空间无法计算和参数的选取问题,研究结果表明,我国制造业增加值逐年增加的趋势还将延续,而制造业增加值占GDP的比重在未来十年内将达到上限值。
Based on the theory of Takens ,in view of the Chinese manufacturing value added accounted for the proportion of GDP , this piece of paper marries together the differential evolution algorithm and support vector machine effectively .This combination integrates model draws the advantage and uses the differential evolution algorithm support vector machine model of China’s manufacturing industry , and the manufacturing value added accounted for the proportion of GDP was predicted based on integration .Simultaneous method can effec-tively avoid the problem of selecting the multidimensional space which cannot be calculated and the parameters of the manufacturing of pre -diction .Results showed that China’s manufacturing industry increased the value of an increasing trend which will continue ,but the manu-facturing value added accounted for the proportion of GDP will reach the upper limit value in the next ten years .
出处
《工业技术经济》
北大核心
2014年第6期116-120,共5页
Journal of Industrial Technological Economics
基金
国家自然科学基金项目"基于自组织的产业结构演化机制与模型研究"(项目编号:71373059)
教育部人文社科基金项目"基于低碳经济的节能减排实现路径与绩效评价研究"(项目编号:13YJA630016)
关键词
微分进化算法
支持向量机
制造业
演化
预测
differential evolution algorithm
support vector machine
manufacturing
evolution
prediction
