Deep4cast: Forecasting for Decision Making under Uncertainty

This package is under active development. Things may change :-).

Deep4Cast is a scalable machine learning package implemented in Python and Torch. It has a front-end API similar to scikit-learn. It is designed for medium to large time series data sets and allows for modeling of forecast uncertainties.

The network architecture is based on WaveNet. Regularization and approximate sampling from posterior predictive distributions of forecasts are achieved via Concrete Dropout.

Documentation is available at read the docs.

Installation

Main Requirements

• python - version 3.6
• pytorch - version 1.0

Source

Before installing we recommend setting up a clean virtual environment.

From the package directory install the requirements and then the package.

$ pip install -r requirements.txt
$ python setup.py install

Examples

• Tutorial Notebooks

Authors:

• Toby Bischoff
• Austin Gross
• Kenneth Tran

References:

• Concrete Dropout is used for approximate posterior Bayesian inference.
• Wavenet is used as encoder network.