This repo aims to be a useful collection of notebooks/code for understanding and implementing seq2seq neural networks for time series forecasting. Networks are constructed with keras/tensorflow.
Navigate to the directory you want the git repo to live in.
- Run
git clone https://github.com/JEddy92/TimeSeries_Seq2Seq.git
- Obtain the wikipedia web traffic data from kaggle. Store it in a folder called "data" at the top level of this repo (this is where the notebooks point to when reading data).