This repository contains the code of [Look Ahead] Improving the Accuracy of Time-Series Forecasting by Previewing Future Time Features
- Python 3.6
- matplotlib == 3.1.1
- numpy == 1.19.4
- pandas == 0.25.1
- scikit_learn == 0.21.3
- torch == 1.8.0
- https://github.com/zhouhaoyi/ETDataset
- https://archive.ics.uci.edu/ml/datasets/Appliances+energy+prediction
- https://archive.ics.uci.edu/ml/datasets/Power+consumption+of+Tetouan+city
In order to train a model for LookAheaed, use test_ax_learning.py script. Following are the main parameters for training:
--len : time stamp embedding input length
--time_point : time stamp after 𝛼 steps
python test_ax_learning.py
python test.py
We appreciate the following github repos a lot for their valuable code base or datasets:
https://github.com/thuml/Autoformer