摘要
为降低经济指标众多及外部因素给经济预测准确度带来的影响,有效提高区域经济预测性能,借助元学习算法的小样本分析优势,提高随机森林算法的适用性,实现区域经济预测。根据区域经济统计数据选取多个数据样本,构建随机森林算法的经济预测模型,通过多个弱分类器投票获得经济预测结果;考虑弱分类器权重数量较少,借助元学习算法对权重进行优化;采用优化的随机森林算法模型完成区域经济预测,并选取中等城市和区不同数量规模的经济样本进行多个经济指标预测仿真。结果表明,经过元学习优化后,随机森林算法在区域生产总值、进口额增长率和居民消费价格指数等方面的预测误差率均有大幅下降,元学习对随机森林算法的优化效果显著。
To reduce the impact of various economic indicators and external factors on the accuracy of economic prediction,regional economic prediction is achieved by taking advantage of small-sample analysis of meta-learning algorithm and the improved random forest algorithm.Multiple data samples are selected from the regional economic statistics to develop an economic prediction model based on random forest algorithm,and the economic prediction results are obtained from multiple weak classifiers voting.As the number of weights of weak classifier is small,meta-learning algorithm is used to optimize the weights.The optimized random forest algorithm economic model is used to predict the regional economy,and different economic samples of medium-sized cities and districts are selected to do simulated prediction of multiple economic indicators.The results show that by optimizing meta-learning,the prediction error rate declines significantly by using random forest algorithm in the economic indicators such as regional GDP,import growth rate and consumer price index,and meta-learning has a remarkable effect on optimization of random forest alogorithm.
作者
李佳颖
吴迪
LI Jia-ying;WU Di(School of Economics and Management,Guangzhou Nanyang Polytechnic College,Guangzhou 510540,China;School of Computer Science and Technology,Harbin Engineering University,Harbin 541004,China;School of Computer and Control Engineering,Qiqihar University,Qiqihar 161006,China)
出处
《南通职业大学学报》
2023年第4期80-85,共6页
Journal of Nantong Vocational University
基金
2021年广东省特色新型智库项目(2021TSZK021)
2022年广东省社科规划项目(GD22XYJ28)。
关键词
经济预测
机器学习
随机森林算法
元学习
模型仿真
权重优化
economic prediction
machine learning
random forest algorithm
meta-learning
model simulation
weight optimization
