youjp / 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 repository implements the common methods of time series prediction, especially deep learning methods in TensorFlow2. It's welcomed to contribute if you have any better idea, just create a PR. If any question, feel free to open an issue.

Ongoing project, I will continue to improve this, so you might want to watch/star this repo to revisit.

RNN	intro	code
wavenet	intro	code
transformer	intro	code
U-Net	intro	code
n-beats	intro	code
GAN	intro	code

Usage

1. Install the required library

pip install -r requirements.txt

2. Download the data, if necessary

bash ./data/download_passenger.sh

3. Train the model
   set custom_model_params if you want (refer to params in ./deepts/models/*.py), and pay attention to feature engineering.

cd examples
python run_train.py --use_model seq2seq
cd ..
tensorboard --logdir=./data/logs

4. Predict new data

cd examples
python run_test.py

Further reading

• https://github.com/awslabs/gluon-ts/
• https://github.com/Azure/DeepLearningForTimeSeriesForecasting
• https://github.com/microsoft/forecasting

Contributor

• LongxingTan