Attention-based novel neural network for mixed frequency data

作     者：Xiangpeng Li Hong Yu Yongfang Xie Jie Li 

作者机构：Chongqing key Laboratory of Computational IntelligenceChongqing University of Posts and TelecommunicationsChongqingChina School of Information Science and EngineeringCentral South UniversityChangshaChina School of Metallurgy and EnvironmentCentral South UniversityChangshaChina 

出 版 物：《CAAI Transactions on Intelligence Technology》 (智能技术学报（英文）)

年 卷 期：2021年第6卷第3期

页      面：301-311页

核心收录：

学科分类：0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 

基　　金：This work was supported in part by the National Natural Science Foundation of China under Grant Nos.61876027,61533020 and 61751312 the key T&A program of Chongqing under grant No.cstc2019jscx-mbdxX0048 

主　　题：network decoder mixed 

摘      要：It is a common fact that data(features,characteristics or variables)are collected at different sampling frequencies in some fields such as economic and *** existing methods usually either ignore the difference from the different sampling frequencies or hardly take notice of the inherent temporal characteristics in mixed frequency *** authors propose an innovative dual attention-based neural network for mixed frequency data(MID-DualAtt),in order to utilize the inherent temporal characteristics and select the input characteristics reasonably without losing *** to the authors’knowledge,this is the first study to use the attention mechanism to process mixed fre-quency *** MID-DualAtt model uses the frequency alignment method to trans-form the high--frequency variables into observation vectors at low frequency,and more critical input characteristics are selected for the current prediction index by attention *** temporal characteristics are explored by the encoder-decoder with attention based on long-short-term memory networks(LSTM).The proposed MID-DualAtt has been tested in practical application,and the experimental results show that it has better prediction ability than the compared models.