Skip to content
Fifth place solution of the Kaggle Diabetic Retinopathy competition.
Jupyter Notebook Python
Branch: master
Clone or download
JeffreyDF Update README.md
Update domain name
Latest commit 77c5781 Jul 12, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data First commit. Aug 2, 2015
dumps Add raw parameter dump and versions in readme. Aug 6, 2015
models Remove temp quick validation for testing. Aug 2, 2015
notebooks Typo fix notebook. Aug 3, 2015
preds First commit. Aug 2, 2015
.gitignore First commit. Aug 2, 2015
LICENSE.md
README.md Update README.md Jul 12, 2019
ds_images_test.py First commit. Aug 2, 2015
ds_images_train.py First commit. Aug 2, 2015
export_params.py Add raw parameter dump and versions in readme. Aug 6, 2015
generators.py Clean up some sys.exit's in generators. Aug 2, 2015
layers.py First commit. Aug 2, 2015
losses.py First commit. Aug 2, 2015
metrics.py First commit. Aug 2, 2015
predict.py First commit. Aug 2, 2015
train.py First commit. Aug 2, 2015
utils.py First commit. Aug 2, 2015

README.md

Kaggle Diabetic Retinopathy Solution.

Fifth place solution for the Kaggle Diabetic Retinopathy competition including some trained network. For more information see my blog post.

The code is quite badly written (lots of ideas, not so much time) but I figured it was more important to release it quickly than to rewrite it all and only publish it weeks later. The most interesting file will probably be the notebook at notebooks/Sample_prediction.ipynb which shows how to load the model dump, do some predictions with it and see the activations for the different layers.

Important to note: the code is built on top of Lasagne at commit cf1a23c21666fc0225a05d284134b255e3613335. For Theano I have been using the latest master and have no problems. Some specific versions:

  • Theano: 9a653e3e91c0e38b6643e4452199931e792a24a2
  • Lasagne: cf1a23c21666fc0225a05d284134b255e3613335
  • Numpy: 1.9.2
  • Pandas: 0.15.2
  • Scikit-learn: 0.16.0
  • Scipy: 0.15.1
  • IPython: 3.0.0
  • Matplotlib: 1.4.2

The basic model included is dependent on cuDNN (but not tested with the latest cuDNN 3 RC). I have, however, also made an export of the raw parameter values for the network using export_params.py. You could then replace the cuDNN layers by the layers of your choice or use these parameters to initialise layers in other frameworks.

You can’t perform that action at this time.