Julian de Wit

Hello.. This is a naming bug from my part. I used to use diameter_mm as a proxy for malignancy.. After that I found the malignancy labels. If I remember correctly...

It's the output. Class is between 0 and 1 (sigmoid). Malignancy is regression between 0 and 25. This is normal convention.

I do not remember if I had that too. It's unusual but it can happen.

Hello.. It's a while ago and I did not use there fields for the model I believe. d = diameter in percentage of image dmm = not used as I...

Hello, I labeled them by hand. They are in the resources file. You don't need this part of the solution to get a good score. It improves everything just a...

Keep it as it is. You can still submit to Kaggle and see the results.

I did local cross validation + cross validation against leaderboard. BOTH needed to show improvements. I tried around 50 features. Almost none gave consistent improvements. Mainly due to the outlier-leaderboard...

It should be in the resources.rar It is generated by training a on LUNA and predicting on luna. Nodules that were false positives were stored.

Hello this constant is used in functions that are not used directly anymore but I kept in for completeness. I'm 99.9% sure that should be : settings.LUNA16_EXTRACTED_IMAGE_DIR

I made a quick and dirty version of the grayscale img iterator. It worked.. no rgb assumptions further in the pipeline.. ... inline static void LoadImage(mshadow::TensorContainer &img, DataInst &out,const char...