- N. Varghese, A. Vitek, Z. Zhao
The training data is large (~10GB uncompressed). It can be downloaded with the following instructions. Alternatively, the iPython notebooks contain saved states with the results.
kaggle competitions download -c petfinder-adoption-prediction
- Ensure the training data is extracted in
./all
- Download https://github.com/git-lfs/git-lfs/releases/download/v2.7.1/git-lfs-darwin-amd64-v2.7.1.tar.gz
git lfs install
git lfs pull
all
contains a single txt file, which will be populated oncegit lfs pull
is performed. Also, it should contain the downloaded training dataattempts
contains the approaches described in the project report. Three are iPython notebooks, one is a regular scriptcheckpoints
contains the best models we obtaineddata_exploration
contains iPython notebooks containing the data analytics we performed, the the Decision Tree discovery in the traditional machine learning pre-trials we made for the project