论文部分内容阅读
国内大豆价格受到多重因素影响,如大豆进口量、国内大豆供给量、中国居民消费价格指数等,因此呈现非线性等特点。大豆价格的剧烈波动会对农户的种植结构和国家政策产生影响,因此准确预测大豆价格具有重要意义。支持向量回归(SVR)因其优越的寻优能力和较高的预测精确度而被广泛应用于非线性时间序列回归中。本文提出一种自适应粒子群算法(APSO)优化的SVR模型来预测我国大豆价格,该模型通过将现实空间内的数据映射到高维空间内,在高维空间内构造线性回归函数,从而判别原有空间内数据之间的关系。在SVR模型参数优化时,由于粒子群算法易陷入局部最优解,因此采用惯性权重更新和适应度变异的粒子群算法(APSO)对预测模型参数进行优化。采用2009年1月-2016年12月的国内大豆价格月度数据进行预测,结果表明APSO优化的SVR模型在大豆价格预测中精度较高,且能准确反应大豆价格的未来趋势,为从事大豆种植者及经营者提供决策依据。
Domestic soybean prices are subject to many factors, such as the import volume of soybeans, the domestic soybean supply, the Chinese consumer price index, etc., and therefore presents a non-linear characteristic. The volatility of soybean prices will have an impact on the planting structure of farmers and national policies. Therefore, it is of great significance to accurately predict the price of soybean. Support Vector Regression (SVR) is widely used in nonlinear time series regression because of its superior searching ability and high prediction accuracy. In this paper, an adaptive particle swarm optimization (APSO) -sized SVR model is proposed to predict the price of soybean in China. This model maps the data in real space into high-dimensional space and constructs linear regression function in high-dimensional space, The relationship between the data in the original space. In the optimization of SVR model parameters, Particle Swarm Optimization (PSSO) is used to optimize the parameters of the prediction model because PSO is easy to fall into the local optimal solution. Based on the monthly data of domestic soybean prices from January 2009 to December 2016, the results show that the APSO-optimized SVR model is more accurate in forecasting soybean prices and can accurately reflect the future trend of soybean prices. And operators to provide decision-making basis.