W2V experiments for CPO & iCIS lecture on bias in ML.
docker create -p 8888:8888 \
-v `pwd`/notebooks:/home/jovyan/work:z \
--name cpo --user `id -u` --group-add users jupyter/minimal-notebook \
start-notebook.sh --NotebookApp.password=`python pwd.py`
docker start cpo
Pretrained W2V vectors are provided by Google: https://code.google.com/archive/p/word2vec/
Download the file GoogleNews-vectors-negative300.bin.gz
.
These pre-trained W2V vectors have also been shared as an Academic Torrent.
sudo dnf install transmission-cli
transmission-cli http://academictorrents.com/download/2aa0d0c6aff92f08719e409db04ecee4721cf21f.torrent
This repository contains three Jupyter notebooks:
- cpo-notebook.ipynb : Notebook used for the CPO & iCIS lecture
- Reproducing_figures_poster_ipynb : Notebook for reproducing all the figures from the poster/paper
- Word2vec_Bias_plots : Experimental notebook, use at own risk.
This work resulted into a short paper and poster presented at FDIA 2019. Both the pdf of the paper and poster can be found here.