Data Science Bowl 2017
Alex Bennet, Anya Gilad, Arpit Vats
Files description:
-
cs640_project_report.pdf - full report
-
rep. Random Forest model -
- luna_preproc.py - creates 3d arrays of the luna dataset
- luna_masks.py - applaying image processing methods to get the masks.
- luna_tf.py - train and learn the luna data set
- random_forest.py - run ML algorithms such as Kmeans and random forest to get a better understaning of the stage1 data set.
-
rep. 3d conv net model - *preproc.py - the 3d image preprocessing we used. *convnet - 3d cnn model with csv output *convnet_exp.py - 3d cnn model with logloss for experiments. *convnet10individuals.py - Experimenting getting the mean probability out of 10 separated experiments (shuffeling the dataset for each run).
-
mergedPrepro.py - New preprocessing method (combining the full preprocessing tutorial and the conv net preprocessing)