LSTM time series forecasting

Car sales forecasting done with LSTM network

Libraries used:

• NumPy
• Pandas
• TensorFlow
• Keras

Virtualenv installation

• Create virtualenv

  $ virtualenv your-venv --no-site-packages

• Activate virtualenv

  $ source your-venv/bin/activate

• Install jupyter in the virtualenv

  (your-venv)$ pip install jupyter

• Add the virtualenv as a jupyter kernel

  (your-venv)$ ipython kernel install --name "your-venv" --user

• List available kernels

  (your-venv)$ jupyter kernelspec list

• Check if kernel.json file in your kernel directory contains correct path to your virtualenv python executable (/path/to/your-env/bin/python3.6)

• You can now select the created kernel your-env when you start Jupyter