FilterNet: A many-to-many deep learning architecture for time series classification
This repository contains code to reproduce the results and figures in the paper: FilterNet: A many-to-many deep learning architecture for time series classification.
The easiest way to run this software is via the Anaconda Python distribution.
- Install Anaconda
conda env create -f environment.yaml
- Enable the
source activate filternet
- Install filternet so it is importable, by running
pip install -e .in the same directory as setup.py
In the root dir of this repo:
This will be really slow the first time because it has to download and pre-process several large AR datasets.
Subsequent test runs will probably still be slow, but... less slow.
Run the scripts in the
scripts/directory. These are very long-running scripts that reproduce each experimental condition many times. You might want to set, e.g.,
NUM_REPEATS=1if you don't need this level of reproducibility.
Run the notebooks to re-produce the figures. You might need to edit a few paths to specific models to match the filenames on your system, especially if you changed the
Copyright (C) 2020 Pet Insight Project - All Rights Reserved