m292m/Time-series-prediction

A collection of time series prediction methods: rnn, seq2seq, cnn, wavenet, transformer, unet, n-beats, gan, kalman-filter

Time series prediction

This repo implements the common methods of time series prediction, especially with deep learning in TensorFlow 2. It's highly welcomed to contribute if you have better idea, just create a PR. If any question, feel free to open an issue.

ARIMA	code
Boosting-tree	code
RNN	code
CNN	code
Transformer	code
U-Net	code
N-Beats	code
GAN	code

Usage

1. Install the library

pip install -r requirements.txt

2. Download the data, if necessary

bash ./data/download_passenger.sh

3. Train the model

cd examples
python run_train.py --use_model seq2seq

set your own model parameters, just set custom_model_params according to each model's params

4. Predict new data

python run_test.py

Further reading

https://github.com/awslabs/gluon-ts/

Contributor

• LongxingTan