提示
基于自适应集成极限学习机的铁路物流景气度预测
首发时间：2024-07-05
摘要：相关物流指数反映了宏观经济运行状况，但不能直观反映铁路物流景气状态。针对小样本数据预测问题，单一极限学习机的预测误差较大，建立针对铁路物流景气度预测的自适应集成极限学习机模型，包括铁路物流景气度定义、预测输入特征体系构建和集成学习框架模型。设计一种自适应集成极限学习机的预测算法，并对置换特征重要性和预测精度进行评估。选择eda数据库近20年的铁路物流数据开展实验，结果表明所建立的预测模型是有效的，且预测精度较次优的未集成的极限学习机综合提高52.76%，较较差的多元线性回归综合提高93.57%。?????
关键词：
for information in english, please click here
prosperity of railway logistics forecasting based on iamelm
abstract：the related logistics index reflects the macroeconomic operating conditions, but cannot intuitively reflect the railway logistics boom state. for the small sample data prediction problem, the prediction error of a single extreme learning machine is large. to establish an integrated adaptive momentum extreme learning machine model for prosperity of railway logistics prediction, including the definition of prosperity of railway logistics, the construction of the prediction input feature system and the integrated learning framework model.design of a prediction algorithm with integrated adaptive momentum extreme learning machine and evaluation of permutation feature importance and prediction accuracy.the railway logistics data of the eda database in the past 20 years were selected to carry out the experiment, and the results showed that the established prediction model was effective,and the prediction accuracy was 52.76% higher than that of the sub-optimal unintegrated extreme learning machine synthesis, and 93.57% higher than that of the poorer multiple linear regression synthesis.
keywords：
论文图表：
引用
导出参考文献
no.****
动态公开评议
共计0人参与
勘误表
基于自适应集成极限学习机的铁路物流景气度预测
评论
全部评论