A temporal-attribute attention neural network for mixed frequency data forecasting(TAA-NN)

Paper title:
A temporal-attribute attention neural network for mixed frequency data forecasting

by Peng Wu and Hong Yu et al.

Copyright (C) 2022
Hong Yu
Chongqing University of Posts and Telecommunications

Usage

• Python 3.6
• Numpy, pandas, functools, sklearn,tensorflow2.0

Describe

The repositories include three folders.

1. “software” folder provides a “.exe” progress allowed reader to reproduce the experiment of mixed-frequency data in this paper.
2. “source code” folder provides the source codes of the TAA-NN algorithm.

Experimental environment:

Pycharm 2020.2.1Edition GPU 1080ti CPU Core i5-8500CPU 3.00GHz Memory RAM16.0GB

Virtual environment:

python 3.6Edition tensorflow 2.2.0Edition keras 2.0.9Edition pandas 0.25.3Edition cython 0.29.14Edition gensim 3.8.3Edition h5py 2.8.0Edition hdf5 1.10.4Edition numpy 1.19.3Edition

reference

Peng Wu, Hong Yu, Feng Hu, Yongfang Xie. A temporal-attribute attention neural network for mixed frequency data forecasting. Int. J. Mach. Learn. & Cyber. 13, 2519–2531 (2022). https://doi.org/10.1007/s13042-022-01541-7