kaggle "MMLM 2022" competition repo.
$ git clone --recursive https://github.com/masafumi330/kaggle_MMLM2022.git
$ sudo apt install cmake build-essential libboost-dev libboost-filesystem-dev
$ cd {your_name}/LightGBM
$ mkdir build && cd build
$ cmake -DUSE_GPU=1 -DOpenCL_LIBRARY=/usr/local/cuda/lib64/libOpenCL.so -DOpenCL_INCLUDE_DIR=/usr/local/cuda/include/ ..
$ make -j7
$ cd ../python-package
$ sudo python setup.py install --precompile --gpu --cuda
pip install --user kaggle
# check if ~/.local/bin exists in $PATH
cd {your_name}/workspace
mkdir data && cd data
kaggle competitions download -c mens-march-mania-2022
unzip mens-march-mania-2022.zip
mkdir -p /workspace/{your_name}/data/ncaa-men-538-team-ratings
cd /workspace/{your_name}/data/ncaa-men-538-team-ratings
touch 538ratingsMen.csv
# write https://www.kaggle.com/raddar/ncaa-men-538-team-ratings?select=538ratingsMen.csv
# into 538ratingsMen.csv
score | issues | who | note |
---|---|---|---|
0.00069 | #15 | Mine | add parameters vol.1 (notebook) |
0.17689 | #12 | Mine | AutoLightGBM with GPU |
0.53075 | #11 | Mine | LightGBM without tuning |
0.54140 | #10 | Mine | not use strategy |
0.56574 | #10 | Mine | safe strategy |
0.65036 | #10 | Mine | risky strategy |